{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "## Perkenalan" ] }, { "cell_type": "code", "execution_count": 334, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "'\\n=================================================\\nGC 5\\n\\nNama : Ryan Trisnadi\\nBatch : HCK-17\\n\\nProyek ini dibuat untuk menguji dan melihat variable dari dataset yang mempunyai efek terbesar terhadap \"default_payment_next_month\" dari dataset \\ncredit_card_default. CSV berisi puluhan kategorikal dan numerikal data yang masing-masing mempunyai faktor terhadap tahap pembayaran utang dari beberapa\\nclient bank/asuransi. \\n\\n\\nPertanyaan:\\n-Resiko apa saja client bank dengan limit balance yang sangat tinggi terhadap utang mereka?\\n-Apakah ada dampak \"education\" terhadap default_payment_next_month?\\n-Adakah perbedaan \"pay\" dan \"bill_amt\" yang bisa disimpulkan terhadap kemampuan client untuk membayar utang balik? \\n-Resiko yang paling tinggi di kolom-kolom apa saja?\\n\\nKita akan menggunakan test-test seperti Logistic Regression, KNN, dan SVM untuk menguji profile client terhadap kemampuan mereka untuk membayar utang \\ndari institusi/bank. \\n\\n=================================================\\n'" ] }, "execution_count": 334, "metadata": {}, "output_type": "execute_result" } ], "source": [ "'''\n", "=================================================\n", "GC 5\n", "\n", "Nama : Ryan Trisnadi\n", "Batch : HCK-17\n", "\n", "Proyek ini dibuat untuk menguji dan melihat variable dari dataset yang mempunyai efek terbesar terhadap \"default_payment_next_month\" dari dataset \n", "credit_card_default. CSV berisi puluhan kategorikal dan numerikal data yang masing-masing mempunyai faktor terhadap tahap pembayaran utang dari beberapa\n", "client bank/asuransi. \n", "\n", "\n", "Pertanyaan:\n", "-Resiko apa saja client bank dengan limit balance yang sangat tinggi terhadap utang mereka?\n", "-Apakah ada dampak \"education\" terhadap default_payment_next_month?\n", "-Adakah perbedaan \"pay\" dan \"bill_amt\" yang bisa disimpulkan terhadap kemampuan client untuk membayar utang balik? \n", "-Resiko yang paling tinggi di kolom-kolom apa saja?\n", "\n", "Kita akan menggunakan test-test seperti Logistic Regression, KNN, dan SVM untuk menguji profile client terhadap kemampuan mereka untuk membayar utang \n", "dari institusi/bank. \n", "\n", "=================================================\n", "'''" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Latar Belakang" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita adalah sebuah analyst yang bekerja di bank di departemen Akuntansi. Tujuan kita untuk menganalisa data-data client yang sudah dan belum bayar tagihan, dan melihat kemampuan mereka untuk membayar tagihan balik dengan tepat waktu. Dengan pola pembayaran mereka setiap interval pembayaran, kita bisa mencoba prediksi kemungkinan mereka akan memenuhi pembayaran utang awal (principal) dan bunga (interest). " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Conceptual Problems\n", "### Jawab pertanyaan berikut:\n", "\n", " Apakah yang dimaksud dengan coeficient pada logistic regression?\n", "\n", " Apakah fungsi parameter kernel pada SVM? Jelaskan salah satu kernel yang kalian pahami!\n", "\n", " Bagaimana cara memilih K yang optimal pada KNN?\n", "\n", " Apa yang dimaksud dengan metrics-metrics berikut : Accuracy, Precision, Recall, F1 Score, dan kapan waktu yang tepat untuk menggunakannya?" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Query SQL" ] }, { "cell_type": "code", "execution_count": 335, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "'\\nfrom google.colab import auth\\nfrom google.cloud import bigquery\\nauth.authenticate_user()\\nprint(\\'Authenticated\\')\\n\\nproject_id = \"elite-outpost-424308-h1\" #GUNAKAN GCP PROJECT-ID KALIAN MASING-MASING\\nclient = bigquery.Client(project=project_id)\\n'" ] }, "execution_count": 335, "metadata": {}, "output_type": "execute_result" } ], "source": [ "\"\"\"\n", "from google.colab import auth\n", "from google.cloud import bigquery\n", "auth.authenticate_user()\n", "print('Authenticated')\n", "\n", "project_id = \"elite-outpost-424308-h1\" #GUNAKAN GCP PROJECT-ID KALIAN MASING-MASING\n", "client = bigquery.Client(project=project_id)\n", "\"\"\"" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Import dari BigQuery lewat Google Colab ke project_id masing-masing. " ] }, { "cell_type": "code", "execution_count": 336, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "\"\\ndf = client.query('''\\nSELECT limit_balance, \\n CAST(sex as INT64) AS sex, \\n CAST(education_level as INT64) AS education_level, \\n CAST(marital_status as INT64) AS marital_status, \\n age, \\n pay_0, \\n pay_2, \\n pay_3, \\n pay_4, \\n CAST(pay_5 AS FLOAT64) AS pay_5, \\n CAST(pay_6 AS FLOAT64) AS pay_6, \\n bill_amt_1, \\n bill_amt_2, \\n bill_amt_3, \\n bill_amt_4, \\n bill_amt_5, \\n bill_amt_6, \\n pay_amt_1, \\n pay_amt_2, \\n pay_amt_3, \\n pay_amt_4, \\n pay_amt_5, \\n pay_amt_6, \\n CAST(default_payment_next_month as INT64) AS default_payment_next_month\\nFROM `bigquery-public-data.ml_datasets.credit_card_default`\\nLIMIT 33966\\n''').to_dataframe()\\n\"" ] }, "execution_count": 336, "metadata": {}, "output_type": "execute_result" } ], "source": [ "\"\"\"\n", "df = client.query('''\n", "SELECT limit_balance, \n", " CAST(sex as INT64) AS sex, \n", " CAST(education_level as INT64) AS education_level, \n", " CAST(marital_status as INT64) AS marital_status, \n", " age, \n", " pay_0, \n", " pay_2, \n", " pay_3, \n", " pay_4, \n", " CAST(pay_5 AS FLOAT64) AS pay_5, \n", " CAST(pay_6 AS FLOAT64) AS pay_6, \n", " bill_amt_1, \n", " bill_amt_2, \n", " bill_amt_3, \n", " bill_amt_4, \n", " bill_amt_5, \n", " bill_amt_6, \n", " pay_amt_1, \n", " pay_amt_2, \n", " pay_amt_3, \n", " pay_amt_4, \n", " pay_amt_5, \n", " pay_amt_6, \n", " CAST(default_payment_next_month as INT64) AS default_payment_next_month\n", "FROM `bigquery-public-data.ml_datasets.credit_card_default`\n", "LIMIT 33966\n", "''').to_dataframe()\n", "\"\"\"" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Waktu import query, kita mau ganti beberapa SELECT function jadi tipe-nya sudah benar. Contohnya: Sex diganti jadi Integer, dan education_level jadi Integer. Ganti tipe untuk lebih mudah mengolah data sebagai numerikal atau kategorikal data. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Huggingface Link" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "https://huggingface.co/spaces/ryantrisnadi/Deployment/tree/main" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Ini link ke website Huggingface dengan data \"Deployment\"." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Import Libraries" ] }, { "cell_type": "code", "execution_count": 337, "metadata": {}, "outputs": [], "source": [ "# Import Library\n", "# Library Dataframe\n", "import pandas as pd\n", "# Library Numerical Data\n", "import numpy as np\n", "# Library Statistic\n", "from scipy import stats\n", "from scipy.stats import uniform\n", "\n", "# Library Data Visualization\n", "import matplotlib.pyplot as plt\n", "import seaborn as sns\n", "\n", "# Library Preprocessing data\n", "from sklearn.model_selection import train_test_split\n", "from sklearn.preprocessing import MinMaxScaler\n", "\n", "# Library Machine Learning Model\n", "from sklearn.metrics.pairwise import manhattan_distances\n", "from sklearn.linear_model import LinearRegression\n", "\n", "import time\n", "from sklearn.neighbors import KNeighborsClassifier\n", "from sklearn.metrics import f1_score, accuracy_score, precision_score, recall_score\n", "from sklearn.linear_model import LogisticRegression\n", "from sklearn.svm import SVC\n", "import matplotlib.pyplot as plt\n", "from sklearn.model_selection import train_test_split\n", "\n", "\n", "# Library Model Evaluation\n", "from sklearn.metrics import ConfusionMatrixDisplay, accuracy_score , classification_report , confusion_matrix, precision_score, recall_score, f1_score\n", "from sklearn.model_selection import RandomizedSearchCV\n", "from sklearn.model_selection import GridSearchCV\n", "from sklearn.model_selection import cross_val_score\n", "from sklearn.model_selection import KFold, StratifiedKFold\n", "from scipy.stats import uniform, randint\n", "\n", "from sklearn.ensemble import RandomForestClassifier\n", "from sklearn.model_selection import RandomizedSearchCV\n", "\n", "# Library Outlier Handling\n", "from feature_engine.outliers import Winsorizer\n", "# Library Correlation\n", "from scipy.stats import kendalltau, pearsonr, spearmanr\n", "\n", "\n", "# Model Evaluation\n", "from sklearn.metrics import mean_squared_error, mean_absolute_error, r2_score\n", "\n", "# Save Model\n", "import pickle\n", "import joblib\n", "import json\n", "\n", "# To Ignore Warning\n", "import warnings\n", "warnings.filterwarnings(\"ignore\")\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita pertama mau impor modul yang akan digunakan untuk menganalisa data dari CSV tersedia. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Data Loading and Cleaning" ] }, { "cell_type": "code", "execution_count": 338, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
limit_balancesexeducation_levelmarital_statusagepay_0pay_2pay_3pay_4pay_5...bill_amt_4bill_amt_5bill_amt_6pay_amt_1pay_amt_2pay_amt_3pay_amt_4pay_amt_5pay_amt_6default_payment_next_month
080000.016154.00.00.00.00.00.0...29296.026210.017643.02545.02208.01336.02232.0542.0348.01
1200000.014149.00.00.00.00.00.0...50146.050235.048984.01689.02164.02500.03480.02500.03000.00
220000.026222.00.00.00.00.00.0...1434.0500.00.04641.01019.0900.00.01500.00.01
3260000.024233.00.00.00.00.00.0...27821.030767.029890.05000.05000.01137.05000.01085.05000.00
4150000.014232.00.00.00.0-1.00.0...150464.0143375.0146411.04019.0146896.0157436.04600.04709.05600.00
..................................................................
296080000.023228.0-1.0-1.0-1.0-2.0-2.0...0.00.00.02800.00.00.00.00.00.00
296150000.023151.0-1.0-1.0-1.0-1.0-2.0...0.00.00.0300.05880.00.00.00.00.01
2962450000.022138.0-2.0-2.0-2.0-2.0-2.0...390.0390.0390.0390.0780.0390.0390.0390.0390.01
296350000.022144.0-2.0-2.0-2.0-2.0-2.0...390.0390.00.0390.0390.0390.0390.00.0780.00
2964290000.022139.01.0-2.0-2.0-2.0-2.0...3184.0390.0390.010000.0800.03184.0390.0390.06617.00
\n", "

2965 rows × 24 columns

\n", "
" ], "text/plain": [ " limit_balance sex education_level marital_status age pay_0 pay_2 \\\n", "0 80000.0 1 6 1 54.0 0.0 0.0 \n", "1 200000.0 1 4 1 49.0 0.0 0.0 \n", "2 20000.0 2 6 2 22.0 0.0 0.0 \n", "3 260000.0 2 4 2 33.0 0.0 0.0 \n", "4 150000.0 1 4 2 32.0 0.0 0.0 \n", "... ... ... ... ... ... ... ... \n", "2960 80000.0 2 3 2 28.0 -1.0 -1.0 \n", "2961 50000.0 2 3 1 51.0 -1.0 -1.0 \n", "2962 450000.0 2 2 1 38.0 -2.0 -2.0 \n", "2963 50000.0 2 2 1 44.0 -2.0 -2.0 \n", "2964 290000.0 2 2 1 39.0 1.0 -2.0 \n", "\n", " pay_3 pay_4 pay_5 ... bill_amt_4 bill_amt_5 bill_amt_6 pay_amt_1 \\\n", "0 0.0 0.0 0.0 ... 29296.0 26210.0 17643.0 2545.0 \n", "1 0.0 0.0 0.0 ... 50146.0 50235.0 48984.0 1689.0 \n", "2 0.0 0.0 0.0 ... 1434.0 500.0 0.0 4641.0 \n", "3 0.0 0.0 0.0 ... 27821.0 30767.0 29890.0 5000.0 \n", "4 0.0 -1.0 0.0 ... 150464.0 143375.0 146411.0 4019.0 \n", "... ... ... ... ... ... ... ... ... \n", "2960 -1.0 -2.0 -2.0 ... 0.0 0.0 0.0 2800.0 \n", "2961 -1.0 -1.0 -2.0 ... 0.0 0.0 0.0 300.0 \n", "2962 -2.0 -2.0 -2.0 ... 390.0 390.0 390.0 390.0 \n", "2963 -2.0 -2.0 -2.0 ... 390.0 390.0 0.0 390.0 \n", "2964 -2.0 -2.0 -2.0 ... 3184.0 390.0 390.0 10000.0 \n", "\n", " pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 pay_amt_6 \\\n", "0 2208.0 1336.0 2232.0 542.0 348.0 \n", "1 2164.0 2500.0 3480.0 2500.0 3000.0 \n", "2 1019.0 900.0 0.0 1500.0 0.0 \n", "3 5000.0 1137.0 5000.0 1085.0 5000.0 \n", "4 146896.0 157436.0 4600.0 4709.0 5600.0 \n", "... ... ... ... ... ... \n", "2960 0.0 0.0 0.0 0.0 0.0 \n", "2961 5880.0 0.0 0.0 0.0 0.0 \n", "2962 780.0 390.0 390.0 390.0 390.0 \n", "2963 390.0 390.0 390.0 0.0 780.0 \n", "2964 800.0 3184.0 390.0 390.0 6617.0 \n", "\n", " default_payment_next_month \n", "0 1 \n", "1 0 \n", "2 1 \n", "3 0 \n", "4 0 \n", "... ... \n", "2960 0 \n", "2961 1 \n", "2962 1 \n", "2963 0 \n", "2964 0 \n", "\n", "[2965 rows x 24 columns]" ] }, "execution_count": 338, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import pandas as pd\n", "\n", "# Replace 'path/to/your/csvfile.csv' with the actual path to your CSV file\n", "file_path = '/Users/ryantrisnadi/Desktop/first_project1/p1-ftds017-hck-g5-ryantrisnadi/_P1G5_Set_1_Ryan_Trisnadi.csv'\n", "\n", "# Read the CSV file into a DataFrame\n", "data = pd.read_csv(file_path)\n", "\n", "# Display the first few rows of the DataFrame\n", "data" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Ini kita extract data CSV tentang credit_card_default dari ml_datasets di BigQuery.\n", "\n", "Legend:\n", "\n", "id = Anonymized ID of each client\n", "\n", "limit_balance = Amount of given credit in NT dollars includes individual and family/supplementary credit\n", "\n", "sex = Gender (1=male, 2=female)\n", "\n", "education_level = Education Level (1=graduate school, 2=university, 3=high school, 4=others, 5=unknown, 6=unknown)\n", "\n", "marital_status = Marital status (1=married, 2=single, 3=others)\n", "\n", "age = Age in years\n", "\n", "pay_0 = Repayment status in September, 2005 (-1=pay duly, 1=payment delay for one month, 2=payment delay for two months, ... 8=payment delay for eight \n", "months, 9=payment delay for nine months and above)\n", "\n", "bill_amt_1 = Amount of bill statement in September, 2005 (NT dollar)\n", "\n", "default_payment_next_month = Default payment (1=yes, 0=no)\n" ] }, { "cell_type": "code", "execution_count": 339, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
limit_balancesexeducation_levelmarital_statusagepay_0pay_2pay_3pay_4pay_5...bill_amt_4bill_amt_5bill_amt_6pay_amt_1pay_amt_2pay_amt_3pay_amt_4pay_amt_5pay_amt_6default_payment_next_month
080000.016154.00.00.00.00.00.0...29296.026210.017643.02545.02208.01336.02232.0542.0348.01
1200000.014149.00.00.00.00.00.0...50146.050235.048984.01689.02164.02500.03480.02500.03000.00
220000.026222.00.00.00.00.00.0...1434.0500.00.04641.01019.0900.00.01500.00.01
3260000.024233.00.00.00.00.00.0...27821.030767.029890.05000.05000.01137.05000.01085.05000.00
4150000.014232.00.00.00.0-1.00.0...150464.0143375.0146411.04019.0146896.0157436.04600.04709.05600.00
5300000.024232.00.00.00.00.00.0...65150.0-450.0700.015235.01491.01303.00.02000.01400.00
6130000.011145.00.00.00.00.00.0...62377.063832.065099.02886.02908.02129.02354.02366.02291.00
7200000.011158.00.00.00.00.00.0...124647.0126921.0129167.07822.04417.04446.04597.04677.04698.00
8500000.011139.00.00.00.00.00.0...174500.0137406.0204975.054209.04607.04603.05224.0207440.07509.00
9230000.011148.00.00.00.00.00.0...105508.0108101.0110094.07000.06607.03773.04290.04164.02000.00
\n", "

10 rows × 24 columns

\n", "
" ], "text/plain": [ " limit_balance sex education_level marital_status age pay_0 pay_2 \\\n", "0 80000.0 1 6 1 54.0 0.0 0.0 \n", "1 200000.0 1 4 1 49.0 0.0 0.0 \n", "2 20000.0 2 6 2 22.0 0.0 0.0 \n", "3 260000.0 2 4 2 33.0 0.0 0.0 \n", "4 150000.0 1 4 2 32.0 0.0 0.0 \n", "5 300000.0 2 4 2 32.0 0.0 0.0 \n", "6 130000.0 1 1 1 45.0 0.0 0.0 \n", "7 200000.0 1 1 1 58.0 0.0 0.0 \n", "8 500000.0 1 1 1 39.0 0.0 0.0 \n", "9 230000.0 1 1 1 48.0 0.0 0.0 \n", "\n", " pay_3 pay_4 pay_5 ... bill_amt_4 bill_amt_5 bill_amt_6 pay_amt_1 \\\n", "0 0.0 0.0 0.0 ... 29296.0 26210.0 17643.0 2545.0 \n", "1 0.0 0.0 0.0 ... 50146.0 50235.0 48984.0 1689.0 \n", "2 0.0 0.0 0.0 ... 1434.0 500.0 0.0 4641.0 \n", "3 0.0 0.0 0.0 ... 27821.0 30767.0 29890.0 5000.0 \n", "4 0.0 -1.0 0.0 ... 150464.0 143375.0 146411.0 4019.0 \n", "5 0.0 0.0 0.0 ... 65150.0 -450.0 700.0 15235.0 \n", "6 0.0 0.0 0.0 ... 62377.0 63832.0 65099.0 2886.0 \n", "7 0.0 0.0 0.0 ... 124647.0 126921.0 129167.0 7822.0 \n", "8 0.0 0.0 0.0 ... 174500.0 137406.0 204975.0 54209.0 \n", "9 0.0 0.0 0.0 ... 105508.0 108101.0 110094.0 7000.0 \n", "\n", " pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 pay_amt_6 \\\n", "0 2208.0 1336.0 2232.0 542.0 348.0 \n", "1 2164.0 2500.0 3480.0 2500.0 3000.0 \n", "2 1019.0 900.0 0.0 1500.0 0.0 \n", "3 5000.0 1137.0 5000.0 1085.0 5000.0 \n", "4 146896.0 157436.0 4600.0 4709.0 5600.0 \n", "5 1491.0 1303.0 0.0 2000.0 1400.0 \n", "6 2908.0 2129.0 2354.0 2366.0 2291.0 \n", "7 4417.0 4446.0 4597.0 4677.0 4698.0 \n", "8 4607.0 4603.0 5224.0 207440.0 7509.0 \n", "9 6607.0 3773.0 4290.0 4164.0 2000.0 \n", "\n", " default_payment_next_month \n", "0 1 \n", "1 0 \n", "2 1 \n", "3 0 \n", "4 0 \n", "5 0 \n", "6 0 \n", "7 0 \n", "8 0 \n", "9 0 \n", "\n", "[10 rows x 24 columns]" ] }, "execution_count": 339, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data.head(10)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Tampil 10 data yang diatas." ] }, { "cell_type": "code", "execution_count": 340, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
limit_balancesexeducation_levelmarital_statusagepay_0pay_2pay_3pay_4pay_5...bill_amt_4bill_amt_5bill_amt_6pay_amt_1pay_amt_2pay_amt_3pay_amt_4pay_amt_5pay_amt_6default_payment_next_month
2955360000.022226.0-1.0-1.0-1.0-1.0-2.0...0.00.00.0463.02500.00.00.00.00.00
2956100000.013140.00.00.0-1.0-1.0-2.0...0.00.00.02000.02377.040000.00.00.00.00
295730000.023148.01.0-1.0-1.0-2.0-2.0...0.00.00.0200.00.00.00.00.00.00
295880000.023139.0-1.0-1.0-1.0-1.0-2.0...0.00.05000.05000.05000.00.05000.05000.0470.00
295920000.013226.0-1.0-1.0-1.0-2.0-2.0...0.00.00.01560.00.00.00.00.00.00
296080000.023228.0-1.0-1.0-1.0-2.0-2.0...0.00.00.02800.00.00.00.00.00.00
296150000.023151.0-1.0-1.0-1.0-1.0-2.0...0.00.00.0300.05880.00.00.00.00.01
2962450000.022138.0-2.0-2.0-2.0-2.0-2.0...390.0390.0390.0390.0780.0390.0390.0390.0390.01
296350000.022144.0-2.0-2.0-2.0-2.0-2.0...390.0390.00.0390.0390.0390.0390.00.0780.00
2964290000.022139.01.0-2.0-2.0-2.0-2.0...3184.0390.0390.010000.0800.03184.0390.0390.06617.00
\n", "

10 rows × 24 columns

\n", "
" ], "text/plain": [ " limit_balance sex education_level marital_status age pay_0 pay_2 \\\n", "2955 360000.0 2 2 2 26.0 -1.0 -1.0 \n", "2956 100000.0 1 3 1 40.0 0.0 0.0 \n", "2957 30000.0 2 3 1 48.0 1.0 -1.0 \n", "2958 80000.0 2 3 1 39.0 -1.0 -1.0 \n", "2959 20000.0 1 3 2 26.0 -1.0 -1.0 \n", "2960 80000.0 2 3 2 28.0 -1.0 -1.0 \n", "2961 50000.0 2 3 1 51.0 -1.0 -1.0 \n", "2962 450000.0 2 2 1 38.0 -2.0 -2.0 \n", "2963 50000.0 2 2 1 44.0 -2.0 -2.0 \n", "2964 290000.0 2 2 1 39.0 1.0 -2.0 \n", "\n", " pay_3 pay_4 pay_5 ... bill_amt_4 bill_amt_5 bill_amt_6 pay_amt_1 \\\n", "2955 -1.0 -1.0 -2.0 ... 0.0 0.0 0.0 463.0 \n", "2956 -1.0 -1.0 -2.0 ... 0.0 0.0 0.0 2000.0 \n", "2957 -1.0 -2.0 -2.0 ... 0.0 0.0 0.0 200.0 \n", "2958 -1.0 -1.0 -2.0 ... 0.0 0.0 5000.0 5000.0 \n", "2959 -1.0 -2.0 -2.0 ... 0.0 0.0 0.0 1560.0 \n", "2960 -1.0 -2.0 -2.0 ... 0.0 0.0 0.0 2800.0 \n", "2961 -1.0 -1.0 -2.0 ... 0.0 0.0 0.0 300.0 \n", "2962 -2.0 -2.0 -2.0 ... 390.0 390.0 390.0 390.0 \n", "2963 -2.0 -2.0 -2.0 ... 390.0 390.0 0.0 390.0 \n", "2964 -2.0 -2.0 -2.0 ... 3184.0 390.0 390.0 10000.0 \n", "\n", " pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 pay_amt_6 \\\n", "2955 2500.0 0.0 0.0 0.0 0.0 \n", "2956 2377.0 40000.0 0.0 0.0 0.0 \n", "2957 0.0 0.0 0.0 0.0 0.0 \n", "2958 5000.0 0.0 5000.0 5000.0 470.0 \n", "2959 0.0 0.0 0.0 0.0 0.0 \n", "2960 0.0 0.0 0.0 0.0 0.0 \n", "2961 5880.0 0.0 0.0 0.0 0.0 \n", "2962 780.0 390.0 390.0 390.0 390.0 \n", "2963 390.0 390.0 390.0 0.0 780.0 \n", "2964 800.0 3184.0 390.0 390.0 6617.0 \n", "\n", " default_payment_next_month \n", "2955 0 \n", "2956 0 \n", "2957 0 \n", "2958 0 \n", "2959 0 \n", "2960 0 \n", "2961 1 \n", "2962 1 \n", "2963 0 \n", "2964 0 \n", "\n", "[10 rows x 24 columns]" ] }, "execution_count": 340, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data.tail(10)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Tampil 10 data yang dibawah." ] }, { "cell_type": "code", "execution_count": 341, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "RangeIndex: 2965 entries, 0 to 2964\n", "Data columns (total 24 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 limit_balance 2965 non-null float64\n", " 1 sex 2965 non-null int64 \n", " 2 education_level 2965 non-null int64 \n", " 3 marital_status 2965 non-null int64 \n", " 4 age 2965 non-null float64\n", " 5 pay_0 2965 non-null float64\n", " 6 pay_2 2965 non-null float64\n", " 7 pay_3 2965 non-null float64\n", " 8 pay_4 2965 non-null float64\n", " 9 pay_5 2965 non-null float64\n", " 10 pay_6 2965 non-null float64\n", " 11 bill_amt_1 2965 non-null float64\n", " 12 bill_amt_2 2965 non-null float64\n", " 13 bill_amt_3 2965 non-null float64\n", " 14 bill_amt_4 2965 non-null float64\n", " 15 bill_amt_5 2965 non-null float64\n", " 16 bill_amt_6 2965 non-null float64\n", " 17 pay_amt_1 2965 non-null float64\n", " 18 pay_amt_2 2965 non-null float64\n", " 19 pay_amt_3 2965 non-null float64\n", " 20 pay_amt_4 2965 non-null float64\n", " 21 pay_amt_5 2965 non-null float64\n", " 22 pay_amt_6 2965 non-null float64\n", " 23 default_payment_next_month 2965 non-null int64 \n", "dtypes: float64(20), int64(4)\n", "memory usage: 556.1 KB\n", "None\n", " limit_balance sex education_level marital_status \\\n", "count 2965.000000 2965.000000 2965.000000 2965.000000 \n", "mean 163369.308600 1.607757 1.849578 1.559865 \n", "std 125030.415472 0.488333 0.778184 0.522317 \n", "min 10000.000000 1.000000 0.000000 0.000000 \n", "25% 50000.000000 1.000000 1.000000 1.000000 \n", "50% 140000.000000 2.000000 2.000000 2.000000 \n", "75% 230000.000000 2.000000 2.000000 2.000000 \n", "max 800000.000000 2.000000 6.000000 3.000000 \n", "\n", " age pay_0 pay_2 pay_3 pay_4 \\\n", "count 2965.000000 2965.000000 2965.000000 2965.000000 2965.000000 \n", "mean 35.193255 0.005059 -0.122428 -0.141653 -0.185160 \n", "std 9.109439 1.114395 1.180784 1.183630 1.178322 \n", "min 21.000000 -2.000000 -2.000000 -2.000000 -2.000000 \n", "25% 28.000000 -1.000000 -1.000000 -1.000000 -1.000000 \n", "50% 34.000000 0.000000 0.000000 0.000000 0.000000 \n", "75% 41.000000 0.000000 0.000000 0.000000 0.000000 \n", "max 69.000000 8.000000 7.000000 7.000000 8.000000 \n", "\n", " pay_5 ... bill_amt_4 bill_amt_5 bill_amt_6 \\\n", "count 2965.000000 ... 2965.000000 2965.000000 2965.000000 \n", "mean -0.225295 ... 44089.683305 40956.080607 39773.072513 \n", "std 1.159003 ... 61907.454056 58271.904751 57303.488981 \n", "min -2.000000 ... -46627.000000 -46627.000000 -73895.000000 \n", "25% -1.000000 ... 2582.000000 1958.000000 1430.000000 \n", "50% 0.000000 ... 19894.000000 18814.000000 18508.000000 \n", "75% 0.000000 ... 58622.000000 53373.000000 52287.000000 \n", "max 7.000000 ... 488808.000000 441981.000000 436172.000000 \n", "\n", " pay_amt_1 pay_amt_2 pay_amt_3 pay_amt_4 \\\n", "count 2965.000000 2.965000e+03 2965.000000 2965.000000 \n", "mean 6348.902867 6.272494e+03 5150.497133 4561.376054 \n", "std 20885.735336 2.887967e+04 14287.079982 13281.499599 \n", "min 0.000000 0.000000e+00 0.000000 0.000000 \n", "25% 1013.000000 9.900000e+02 477.000000 313.000000 \n", "50% 2234.000000 2.175000e+03 1994.000000 1600.000000 \n", "75% 5087.000000 5.000000e+03 4500.000000 4000.000000 \n", "max 493358.000000 1.227082e+06 199209.000000 202076.000000 \n", "\n", " pay_amt_5 pay_amt_6 default_payment_next_month \n", "count 2965.000000 2965.000000 2965.000000 \n", "mean 4913.286678 5382.701518 0.214165 \n", "std 16734.340778 17275.953029 0.410311 \n", "min 0.000000 0.000000 0.000000 \n", "25% 323.000000 173.000000 0.000000 \n", "50% 1646.000000 1615.000000 0.000000 \n", "75% 4021.000000 4081.000000 0.000000 \n", "max 388071.000000 403500.000000 1.000000 \n", "\n", "[8 rows x 24 columns]\n" ] } ], "source": [ "# Check the structure and basic statistics of the data\n", "print(data.info())\n", "print(data.describe())" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Ini untuk chek tipe dari semua kolom di data CSV. " ] }, { "cell_type": "code", "execution_count": 342, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Index(['limit_balance', 'sex', 'education_level', 'marital_status', 'age',\n", " 'pay_0', 'pay_2', 'pay_3', 'pay_4', 'pay_5', 'pay_6', 'bill_amt_1',\n", " 'bill_amt_2', 'bill_amt_3', 'bill_amt_4', 'bill_amt_5', 'bill_amt_6',\n", " 'pay_amt_1', 'pay_amt_2', 'pay_amt_3', 'pay_amt_4', 'pay_amt_5',\n", " 'pay_amt_6', 'default_payment_next_month'],\n", " dtype='object')" ] }, "execution_count": 342, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data.columns" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Lihat semua kolom dari table tersebut." ] }, { "cell_type": "code", "execution_count": 343, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
limit_balancesexeducation_levelmarital_statusagepay_0pay_2pay_3pay_4pay_5...bill_amt_4bill_amt_5bill_amt_6pay_amt_1pay_amt_2pay_amt_3pay_amt_4pay_amt_5pay_amt_6default_payment_next_month
080000.016154.00.00.00.00.00.0...29296.026210.017643.02545.02208.01336.02232.0542.0348.01
1200000.014149.00.00.00.00.00.0...50146.050235.048984.01689.02164.02500.03480.02500.03000.00
220000.026222.00.00.00.00.00.0...1434.0500.00.04641.01019.0900.00.01500.00.01
3260000.024233.00.00.00.00.00.0...27821.030767.029890.05000.05000.01137.05000.01085.05000.00
4150000.014232.00.00.00.0-1.00.0...150464.0143375.0146411.04019.0146896.0157436.04600.04709.05600.00
..................................................................
296080000.023228.0-1.0-1.0-1.0-2.0-2.0...0.00.00.02800.00.00.00.00.00.00
296150000.023151.0-1.0-1.0-1.0-1.0-2.0...0.00.00.0300.05880.00.00.00.00.01
2962450000.022138.0-2.0-2.0-2.0-2.0-2.0...390.0390.0390.0390.0780.0390.0390.0390.0390.01
296350000.022144.0-2.0-2.0-2.0-2.0-2.0...390.0390.00.0390.0390.0390.0390.00.0780.00
2964290000.022139.01.0-2.0-2.0-2.0-2.0...3184.0390.0390.010000.0800.03184.0390.0390.06617.00
\n", "

2964 rows × 24 columns

\n", "
" ], "text/plain": [ " limit_balance sex education_level marital_status age pay_0 pay_2 \\\n", "0 80000.0 1 6 1 54.0 0.0 0.0 \n", "1 200000.0 1 4 1 49.0 0.0 0.0 \n", "2 20000.0 2 6 2 22.0 0.0 0.0 \n", "3 260000.0 2 4 2 33.0 0.0 0.0 \n", "4 150000.0 1 4 2 32.0 0.0 0.0 \n", "... ... ... ... ... ... ... ... \n", "2960 80000.0 2 3 2 28.0 -1.0 -1.0 \n", "2961 50000.0 2 3 1 51.0 -1.0 -1.0 \n", "2962 450000.0 2 2 1 38.0 -2.0 -2.0 \n", "2963 50000.0 2 2 1 44.0 -2.0 -2.0 \n", "2964 290000.0 2 2 1 39.0 1.0 -2.0 \n", "\n", " pay_3 pay_4 pay_5 ... bill_amt_4 bill_amt_5 bill_amt_6 pay_amt_1 \\\n", "0 0.0 0.0 0.0 ... 29296.0 26210.0 17643.0 2545.0 \n", "1 0.0 0.0 0.0 ... 50146.0 50235.0 48984.0 1689.0 \n", "2 0.0 0.0 0.0 ... 1434.0 500.0 0.0 4641.0 \n", "3 0.0 0.0 0.0 ... 27821.0 30767.0 29890.0 5000.0 \n", "4 0.0 -1.0 0.0 ... 150464.0 143375.0 146411.0 4019.0 \n", "... ... ... ... ... ... ... ... ... \n", "2960 -1.0 -2.0 -2.0 ... 0.0 0.0 0.0 2800.0 \n", "2961 -1.0 -1.0 -2.0 ... 0.0 0.0 0.0 300.0 \n", "2962 -2.0 -2.0 -2.0 ... 390.0 390.0 390.0 390.0 \n", "2963 -2.0 -2.0 -2.0 ... 390.0 390.0 0.0 390.0 \n", "2964 -2.0 -2.0 -2.0 ... 3184.0 390.0 390.0 10000.0 \n", "\n", " pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 pay_amt_6 \\\n", "0 2208.0 1336.0 2232.0 542.0 348.0 \n", "1 2164.0 2500.0 3480.0 2500.0 3000.0 \n", "2 1019.0 900.0 0.0 1500.0 0.0 \n", "3 5000.0 1137.0 5000.0 1085.0 5000.0 \n", "4 146896.0 157436.0 4600.0 4709.0 5600.0 \n", "... ... ... ... ... ... \n", "2960 0.0 0.0 0.0 0.0 0.0 \n", "2961 5880.0 0.0 0.0 0.0 0.0 \n", "2962 780.0 390.0 390.0 390.0 390.0 \n", "2963 390.0 390.0 390.0 0.0 780.0 \n", "2964 800.0 3184.0 390.0 390.0 6617.0 \n", "\n", " default_payment_next_month \n", "0 1 \n", "1 0 \n", "2 1 \n", "3 0 \n", "4 0 \n", "... ... \n", "2960 0 \n", "2961 1 \n", "2962 1 \n", "2963 0 \n", "2964 0 \n", "\n", "[2964 rows x 24 columns]" ] }, "execution_count": 343, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data.drop_duplicates()" ] }, { "cell_type": "code", "execution_count": 344, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "limit_balance 0\n", "sex 0\n", "education_level 0\n", "marital_status 0\n", "age 0\n", "pay_0 0\n", "pay_2 0\n", "pay_3 0\n", "pay_4 0\n", "pay_5 0\n", "pay_6 0\n", "bill_amt_1 0\n", "bill_amt_2 0\n", "bill_amt_3 0\n", "bill_amt_4 0\n", "bill_amt_5 0\n", "bill_amt_6 0\n", "pay_amt_1 0\n", "pay_amt_2 0\n", "pay_amt_3 0\n", "pay_amt_4 0\n", "pay_amt_5 0\n", "pay_amt_6 0\n", "default_payment_next_month 0\n", "dtype: int64\n" ] } ], "source": [ "# Check for missing values\n", "print(data.isnull().sum())" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita mau lihat jika ada data yang null atau kosong. Kelihatan semua kolom termasuk values dan terisi." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Exploratory Data Analysis" ] }, { "cell_type": "code", "execution_count": 345, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
limit_balancesexeducation_levelmarital_statusagepay_0pay_2pay_3pay_4pay_5...bill_amt_4bill_amt_5bill_amt_6pay_amt_1pay_amt_2pay_amt_3pay_amt_4pay_amt_5pay_amt_6default_payment_next_month
1200000.014149.00.00.00.00.00.0...50146.050235.048984.01689.02164.02500.03480.02500.03000.00
3260000.024233.00.00.00.00.00.0...27821.030767.029890.05000.05000.01137.05000.01085.05000.00
4150000.014232.00.00.00.0-1.00.0...150464.0143375.0146411.04019.0146896.0157436.04600.04709.05600.00
5300000.024232.00.00.00.00.00.0...65150.0-450.0700.015235.01491.01303.00.02000.01400.00
6130000.011145.00.00.00.00.00.0...62377.063832.065099.02886.02908.02129.02354.02366.02291.00
..................................................................
295880000.023139.0-1.0-1.0-1.0-1.0-2.0...0.00.05000.05000.05000.00.05000.05000.0470.00
295920000.013226.0-1.0-1.0-1.0-2.0-2.0...0.00.00.01560.00.00.00.00.00.00
296080000.023228.0-1.0-1.0-1.0-2.0-2.0...0.00.00.02800.00.00.00.00.00.00
296350000.022144.0-2.0-2.0-2.0-2.0-2.0...390.0390.00.0390.0390.0390.0390.00.0780.00
2964290000.022139.01.0-2.0-2.0-2.0-2.0...3184.0390.0390.010000.0800.03184.0390.0390.06617.00
\n", "

2330 rows × 24 columns

\n", "
" ], "text/plain": [ " limit_balance sex education_level marital_status age pay_0 pay_2 \\\n", "1 200000.0 1 4 1 49.0 0.0 0.0 \n", "3 260000.0 2 4 2 33.0 0.0 0.0 \n", "4 150000.0 1 4 2 32.0 0.0 0.0 \n", "5 300000.0 2 4 2 32.0 0.0 0.0 \n", "6 130000.0 1 1 1 45.0 0.0 0.0 \n", "... ... ... ... ... ... ... ... \n", "2958 80000.0 2 3 1 39.0 -1.0 -1.0 \n", "2959 20000.0 1 3 2 26.0 -1.0 -1.0 \n", "2960 80000.0 2 3 2 28.0 -1.0 -1.0 \n", "2963 50000.0 2 2 1 44.0 -2.0 -2.0 \n", "2964 290000.0 2 2 1 39.0 1.0 -2.0 \n", "\n", " pay_3 pay_4 pay_5 ... bill_amt_4 bill_amt_5 bill_amt_6 pay_amt_1 \\\n", "1 0.0 0.0 0.0 ... 50146.0 50235.0 48984.0 1689.0 \n", "3 0.0 0.0 0.0 ... 27821.0 30767.0 29890.0 5000.0 \n", "4 0.0 -1.0 0.0 ... 150464.0 143375.0 146411.0 4019.0 \n", "5 0.0 0.0 0.0 ... 65150.0 -450.0 700.0 15235.0 \n", "6 0.0 0.0 0.0 ... 62377.0 63832.0 65099.0 2886.0 \n", "... ... ... ... ... ... ... ... ... \n", "2958 -1.0 -1.0 -2.0 ... 0.0 0.0 5000.0 5000.0 \n", "2959 -1.0 -2.0 -2.0 ... 0.0 0.0 0.0 1560.0 \n", "2960 -1.0 -2.0 -2.0 ... 0.0 0.0 0.0 2800.0 \n", "2963 -2.0 -2.0 -2.0 ... 390.0 390.0 0.0 390.0 \n", "2964 -2.0 -2.0 -2.0 ... 3184.0 390.0 390.0 10000.0 \n", "\n", " pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 pay_amt_6 \\\n", "1 2164.0 2500.0 3480.0 2500.0 3000.0 \n", "3 5000.0 1137.0 5000.0 1085.0 5000.0 \n", "4 146896.0 157436.0 4600.0 4709.0 5600.0 \n", "5 1491.0 1303.0 0.0 2000.0 1400.0 \n", "6 2908.0 2129.0 2354.0 2366.0 2291.0 \n", "... ... ... ... ... ... \n", "2958 5000.0 0.0 5000.0 5000.0 470.0 \n", "2959 0.0 0.0 0.0 0.0 0.0 \n", "2960 0.0 0.0 0.0 0.0 0.0 \n", "2963 390.0 390.0 390.0 0.0 780.0 \n", "2964 800.0 3184.0 390.0 390.0 6617.0 \n", "\n", " default_payment_next_month \n", "1 0 \n", "3 0 \n", "4 0 \n", "5 0 \n", "6 0 \n", "... ... \n", "2958 0 \n", "2959 0 \n", "2960 0 \n", "2963 0 \n", "2964 0 \n", "\n", "[2330 rows x 24 columns]" ] }, "execution_count": 345, "metadata": {}, "output_type": "execute_result" } ], "source": [ "group_0 = data[data['default_payment_next_month'] == 0]\n", "group_0" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Disini, kita mau coba group data dengan yang nilai \"0\" di \"default_payment_next_month\". Ini adalah clients yang masih lunas dalam tahap pembayaran mereka dan belum \"default\". " ] }, { "cell_type": "code", "execution_count": 346, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
limit_balancesexeducation_levelmarital_statusagepay_0pay_2pay_3pay_4pay_5...bill_amt_4bill_amt_5bill_amt_6pay_amt_1pay_amt_2pay_amt_3pay_amt_4pay_amt_5pay_amt_6default_payment_next_month
080000.016154.00.00.00.00.00.0...29296.026210.017643.02545.02208.01336.02232.0542.0348.01
220000.026222.00.00.00.00.00.0...1434.0500.00.04641.01019.0900.00.01500.00.01
18360000.011146.00.00.00.00.00.0...13780.015077.014009.03005.03024.04004.04008.03008.02024.01
5220000.011222.00.00.00.00.00.0...22231.022301.021687.01331.01675.01327.0800.0783.0777.01
59100000.011230.00.00.00.00.00.0...97862.079099.079812.04511.03711.03685.02797.02897.03046.01
..................................................................
2942430000.012132.01.0-1.0-1.0-2.0-2.0...0.00.00.02500.00.00.00.00.00.01
294420000.022138.01.0-1.0-1.0-2.0-2.0...0.00.00.02000.00.00.00.00.00.01
295210000.012130.0-1.0-1.0-1.0-1.0-2.0...0.00.00.00.0780.00.00.00.00.01
296150000.023151.0-1.0-1.0-1.0-1.0-2.0...0.00.00.0300.05880.00.00.00.00.01
2962450000.022138.0-2.0-2.0-2.0-2.0-2.0...390.0390.0390.0390.0780.0390.0390.0390.0390.01
\n", "

635 rows × 24 columns

\n", "
" ], "text/plain": [ " limit_balance sex education_level marital_status age pay_0 pay_2 \\\n", "0 80000.0 1 6 1 54.0 0.0 0.0 \n", "2 20000.0 2 6 2 22.0 0.0 0.0 \n", "18 360000.0 1 1 1 46.0 0.0 0.0 \n", "52 20000.0 1 1 2 22.0 0.0 0.0 \n", "59 100000.0 1 1 2 30.0 0.0 0.0 \n", "... ... ... ... ... ... ... ... \n", "2942 430000.0 1 2 1 32.0 1.0 -1.0 \n", "2944 20000.0 2 2 1 38.0 1.0 -1.0 \n", "2952 10000.0 1 2 1 30.0 -1.0 -1.0 \n", "2961 50000.0 2 3 1 51.0 -1.0 -1.0 \n", "2962 450000.0 2 2 1 38.0 -2.0 -2.0 \n", "\n", " pay_3 pay_4 pay_5 ... bill_amt_4 bill_amt_5 bill_amt_6 pay_amt_1 \\\n", "0 0.0 0.0 0.0 ... 29296.0 26210.0 17643.0 2545.0 \n", "2 0.0 0.0 0.0 ... 1434.0 500.0 0.0 4641.0 \n", "18 0.0 0.0 0.0 ... 13780.0 15077.0 14009.0 3005.0 \n", "52 0.0 0.0 0.0 ... 22231.0 22301.0 21687.0 1331.0 \n", "59 0.0 0.0 0.0 ... 97862.0 79099.0 79812.0 4511.0 \n", "... ... ... ... ... ... ... ... ... \n", "2942 -1.0 -2.0 -2.0 ... 0.0 0.0 0.0 2500.0 \n", "2944 -1.0 -2.0 -2.0 ... 0.0 0.0 0.0 2000.0 \n", "2952 -1.0 -1.0 -2.0 ... 0.0 0.0 0.0 0.0 \n", "2961 -1.0 -1.0 -2.0 ... 0.0 0.0 0.0 300.0 \n", "2962 -2.0 -2.0 -2.0 ... 390.0 390.0 390.0 390.0 \n", "\n", " pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 pay_amt_6 \\\n", "0 2208.0 1336.0 2232.0 542.0 348.0 \n", "2 1019.0 900.0 0.0 1500.0 0.0 \n", "18 3024.0 4004.0 4008.0 3008.0 2024.0 \n", "52 1675.0 1327.0 800.0 783.0 777.0 \n", "59 3711.0 3685.0 2797.0 2897.0 3046.0 \n", "... ... ... ... ... ... \n", "2942 0.0 0.0 0.0 0.0 0.0 \n", "2944 0.0 0.0 0.0 0.0 0.0 \n", "2952 780.0 0.0 0.0 0.0 0.0 \n", "2961 5880.0 0.0 0.0 0.0 0.0 \n", "2962 780.0 390.0 390.0 390.0 390.0 \n", "\n", " default_payment_next_month \n", "0 1 \n", "2 1 \n", "18 1 \n", "52 1 \n", "59 1 \n", "... ... \n", "2942 1 \n", "2944 1 \n", "2952 1 \n", "2961 1 \n", "2962 1 \n", "\n", "[635 rows x 24 columns]" ] }, "execution_count": 346, "metadata": {}, "output_type": "execute_result" } ], "source": [ "group_1 = data[data['default_payment_next_month'] == 1]\n", "group_1" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Disini, kita mau coba group data dengan yang nilai \"1\" di \"default_payment_next_month\". Ini adalah clients yang sudah \"default\" dan harus ditagih untuk pembayaran yang terlambat. " ] }, { "cell_type": "code", "execution_count": 347, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([1, 2])" ] }, "execution_count": 347, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data['sex'].unique()" ] }, { "cell_type": "code", "execution_count": 348, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([1, 2, 3, 0])" ] }, "execution_count": 348, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data['marital_status'].unique()" ] }, { "cell_type": "code", "execution_count": 349, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([6, 4, 1, 2, 3, 5, 0])" ] }, "execution_count": 349, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data['education_level'].unique()" ] }, { "cell_type": "code", "execution_count": 350, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([54., 49., 22., 33., 32., 45., 58., 39., 48., 34., 47., 46., 30.,\n", " 35., 55., 42., 56., 31., 53., 40., 36., 51., 37., 44., 24., 38.,\n", " 26., 25., 23., 27., 28., 29., 41., 63., 50., 43., 66., 61., 52.,\n", " 62., 69., 21., 65., 57., 64., 67., 60., 59., 68.])" ] }, "execution_count": 350, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data['age'].unique()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita mau cek unique values dari beberapa kolom untuk melihat cardinality/variance dari setiap values. " ] }, { "cell_type": "code", "execution_count": 351, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sns.countplot(x='default_payment_next_month', data=data)\n", "plt.title('Distribution of Default Payment Next Month')\n", "plt.xlabel('Default Payment Next Month')\n", "plt.ylabel('Count')\n", "plt.show()\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita mau cek beberapa client yang lulus bayar utang \"0\", dan yang gagal bayar utang \"1\" dari di semua dataset. " ] }, { "cell_type": "code", "execution_count": 352, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sns.histplot(data=data, x='age', hue='default_payment_next_month', bins=20, kde=True)\n", "plt.title('Distribution of Age by Default Payment Next Month')\n", "plt.xlabel('Age')\n", "plt.ylabel('Frequency')\n", "plt.show()\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Juga mau cek distribusi dari semua client yang lunas atau gagal bayar utang dari keseluruhan dataset. Dari distribusi yang skewed positive, client cenderung kurang dari umur 40-an daripada yang lebih tua dari 40 tahun. " ] }, { "cell_type": "code", "execution_count": 353, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sns.boxplot(x='default_payment_next_month', y='limit_balance', data=data)\n", "plt.title('Limit Balance by Default Payment Next Month')\n", "plt.xlabel('Default Payment Next Month')\n", "plt.ylabel('Limit Balance')\n", "plt.show()\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Disini, kita mau lihat perbedaan client yang lunas dan gagal bayar di variable limit balance dengan box plot. Kelihatan client yang lunas bayar tagihan memiliki Limit_balance yang lebih tinggi daripada yang gagal bayar. Outlier untuk client yang mampu bayar tagihan juga lebih luas daripada yang gagal bayar. Ini artinya client-client yang tidak pernah gagal payar lebih mampu untuk ambil pinjaman dari bank dengan balance yang lebih tinggi, juga kemungkinan besar mereka ditawarkan \"loan\" dengan quota lebih besar setiap bulan. " ] }, { "cell_type": "code", "execution_count": 354, "metadata": {}, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkQAAAHHCAYAAABeLEexAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy80BEi2AAAACXBIWXMAAA9hAAAPYQGoP6dpAABSQUlEQVR4nO3deVgVZf8G8Puw74dFVmMTN1DcQAn3hUQjX7fX1FDRzC1xo6wsBZfSFFNTCcM30QpyqTQ1xRTXFDeUXFNUTFMBN0BQWZ/fH17MzyOggMA5OPfnus51cWaeM/Odh4Fzn5ln5iiEEAJEREREMqal7gKIiIiI1I2BiIiIiGSPgYiIiIhkj4GIiIiIZI+BiIiIiGSPgYiIiIhkj4GIiIiIZI+BiIiIiGSPgYiIiIhkj4GIyuXq1atQKBRYvXq1uksp0969e6FQKLB37151l6JRVq9eDYVCgatXrz633fDhw2FiYlIzRVWj8PBw1KtXD9ra2mjRokW1rIP7Gr1Kivfnn3/+Wd2lqBUDkUwUvymW9Th8+LC6Syy3b775RuOCWefOndG0aVN1l1EruLi4SPudlpYWzM3N4enpidGjR+PIkSMvtew//vgDH330Edq1a4fo6GjMnTu3iqp+sdjYWCxZsqTc7Z/uB4VCARsbG3To0AEbN26sviI10LZt2zBz5sxyt+/cuTMUCgV69epVYl7xB7eFCxdWYYWqDh06hJkzZyIjI6Nc7YcPHw6FQgEzMzM8evSoxPzk5GRpH6jOuoGK76Nyo6PuAqhmzZ49G66uriWm169fXw3VVM4333yDOnXqYPjw4SrTO3bsiEePHkFPT089hVG5tWjRAh988AEA4MGDBzh//jw2bNiAlStXYsqUKVi0aFGllrt7925oaWnhu+++q/H9IDY2FmfOnMHkyZPL/Zqn++HmzZv49ttv0a9fP0RGRmLs2LHVVKlm2bZtGyIiIioUigBg69atSExMhJeXV/UUVoZDhw5h1qxZGD58OMzNzcv1Gh0dHTx8+BBbtmzB22+/rTIvJiYGBgYGePz4cTVUq6oy+6icMBDJTM+ePeHt7a3uMqqFlpYWDAwM1F0GlUPdunUxZMgQlWnz58/HO++8g8WLF6NBgwYYN25chZebnp4OQ0PDWhOKn+2HYcOGoX79+li8eLFsAlFlODk54cGDB5g1axY2b96s7nJeSF9fH+3atcNPP/1UIhDFxsYiICAAv/zyi5qqo2I8ZUYlZGRkYPjw4VAqlTA3N0dQUFCph4c7d+6Mzp07l5g+fPhwuLi4qEwrKirC119/DU9PTxgYGMDa2ho9evTA8ePHpTbR0dHo2rUrbGxsoK+vDw8PD0RGRqosx8XFBWfPnsW+ffukw8zFNZQ1rmPDhg3w8vKCoaEh6tSpgyFDhuDGjRslajYxMcGNGzfQp08fmJiYwNraGh9++CEKCwvL3Xcvsn37dnTo0AHGxsYwNTVFQEAAzp49K81fuHAhFAoF/vnnnxKvnTZtGvT09HD//n1p2pEjR9CjRw8olUoYGRmhU6dOOHjw4EvVeOXKFfj7+8PY2BgODg6YPXs2hBAAACEEXFxc0Lt37xKve/z4MZRKJcaMGVOp9RoaGuKHH36ApaUlvvjiC2mdwJP9Z8mSJWjSpAkMDAxga2uLMWPGqPSFQqFAdHQ0cnJypH2j+NRqefat4mWUdqTCxcWlxBHJp3Xu3Bm///47/vnnH2ndz/4NlIednR3c3d2RkpICADh16hSGDx+OevXqwcDAAHZ2dnj33Xdx9+5d6TV79uyBQqEo9VRbbGwsFAoFEhISAPz/fn7t2jW89dZbMDExQd26dREREQEAOH36NLp27QpjY2M4OzsjNja2xDIzMjIwefJkODo6Ql9fH/Xr18f8+fNRVFQktXn61FVUVBTc3Nygr6+P1q1b49ixY1K74cOHS+t++vThi5iammLKlCnYsmULTpw48cL2L6pZCIEuXbrA2toa6enp0uvy8vLg6ekJNzc35OTkYObMmZg6dSoAwNXVVar3RePzAOCdd97B9u3bVf6XHjt2DMnJyXjnnXdKfc2VK1cwYMAAWFpawsjICK+//jp+//13lTbF//fWr1+PL774Aq+99hoMDAzQrVs3XLp0SWpXnn20qKjouct41fEIkcxkZmbizp07KtMUCgWsrKwAPPnH0Lt3b/z5558YO3Ys3N3dsXHjRgQFBb3UekeOHInVq1ejZ8+eeO+991BQUIADBw7g8OHD0hGryMhINGnSBP/5z3+go6ODLVu24P3330dRURHGjx8PAFiyZAkmTJgAExMTfPbZZwAAW1vbMte7evVqjBgxAq1bt8a8efOQlpaGr7/+GgcPHsTJkydVDnkXFhbC398fPj4+WLhwIXbt2oWvvvoKbm5ulTpa8awffvgBQUFB8Pf3x/z58/Hw4UNERkaiffv2OHnyJFxcXPD222/jo48+wvr166V/vMXWr1+P7t27w8LCAsCT00M9e/aEl5cXwsLCoKWlJb3xHzhwAG3atKlwjYWFhejRowdef/11LFiwAHFxcQgLC0NBQQFmz54NhUKBIUOGYMGCBbh37x4sLS2l127ZsgVZWVkljvxUhImJCfr27YvvvvsO586dQ5MmTQAAY8aMkX6XEydOREpKCpYvX46TJ0/i4MGD0NXVxQ8//ICoqCgcPXoU//vf/wAAbdu2BVC+fetlfPbZZ8jMzMS///6LxYsXS9tSUfn5+bh+/br097hz505cuXIFI0aMgJ2dHc6ePYuoqCicPXsWhw8flj4QODo6IiYmBn379lVZXkxMDNzc3ODr6ytNKywsRM+ePdGxY0csWLAAMTExCA4OhrGxMT777DMEBgaiX79+WLFiBYYNGwZfX1/pNPvDhw/RqVMn3LhxA2PGjIGTkxMOHTqEadOm4datWyXGp8TGxuLBgwcYM2YMFAoFFixYgH79+uHKlSvQ1dXFmDFjcPPmTezcuRM//PBDhfpq0qRJWLx4MWbOnPnco0TlqVmhUGDVqlVo1qwZxo4di19//RUAEBYWhrNnz2Lv3r0wNjZGv379cPHiRfz0009YvHgx6tSpAwCwtrZ+Yb39+vWTlv3uu+9K/dO4cWO0atWqRPu0tDS0bdsWDx8+xMSJE2FlZYU1a9bgP//5D37++ecSv+svv/wSWlpa+PDDD5GZmYkFCxYgMDBQGpdXnn30Rct45QmShejoaAGg1Ie+vr7UbtOmTQKAWLBggTStoKBAdOjQQQAQ0dHR0vROnTqJTp06lVhXUFCQcHZ2lp7v3r1bABATJ04s0baoqEj6+eHDhyXm+/v7i3r16qlMa9KkSanr3bNnjwAg9uzZI4QQIi8vT9jY2IimTZuKR48eSe22bt0qAIjQ0FCVmgGI2bNnqyyzZcuWwsvLq8S6ntWpUyfRpEmTMuc/ePBAmJubi1GjRqlMT01NFUqlUmW6r69viXUePXpUABDff/+9EOJJvzVo0ED4+/uX6ENXV1fxxhtvSNOKf/cpKSnP3YbiPpgwYYI0raioSAQEBAg9PT1x+/ZtIYQQFy5cEABEZGSkyuv/85//CBcXF5V6SuPs7CwCAgLKnL948WIBQPz2229CCCEOHDggAIiYmBiVdnFxcSWmBwUFCWNj4xLLLO++BUCEhYWVWnNQUJD0/Nl9TQghAgICVPb7F3F2dhbdu3cXt2/fFrdv3xZ//fWXGDRokMrvoLS6f/rpJwFA7N+/X5o2bdo0oa+vLzIyMqRp6enpQkdHR2V7in/Hc+fOlabdv39fGBoaCoVCIdauXStN//vvv0v0x5w5c4SxsbG4ePGiSk2ffPKJ0NbWFteuXRNCCJGSkiIACCsrK3Hv3j2p3W+//SYAiC1btkjTxo8fLyryVvT039qsWbMEAJGYmKiy3vDw8ArXLIQQ3377rQAgfvzxR3H48GGhra0tJk+erPK68PDwcv09FXt6n/zvf/8runXrJoQQorCwUNjZ2YlZs2aVWvfkyZMFAHHgwAFp2oMHD4Srq6twcXERhYWFQoj/3xfd3d1Fbm6u1Pbrr78WAMTp06elaWXtoxVZxquMp8xkJiIiAjt37lR5bN++XZq/bds26OjoqBwR0dbWxoQJEyq9zl9++QUKhQJhYWEl5j19eNzQ0FD6ufhIVqdOnXDlyhVkZmZWeL3Hjx9Heno63n//fZWxRQEBAWjcuHGJQ88ASozb6NChA65cuVLhdT9r586dyMjIwODBg3Hnzh3poa2tDR8fH+zZs0dqO3DgQCQmJuLy5cvStHXr1kFfX186VZWUlCQdar979660vJycHHTr1g379+9XOYVREcHBwdLPCoUCwcHByMvLw65duwAADRs2hI+PD2JiYqR29+7dw/bt2xEYGFiuUx7PU/yp9cGDBwCenPJUKpV44403VPrOy8sLJiYmKn1Xlqret6rKH3/8AWtra1hbW6N58+bYsGEDhg4divnz5wNQrfvx48e4c+cOXn/9dQBQOVU0bNgw5Obmqlw2vW7dOhQUFJR6xO69996TfjY3N0ejRo1gbGysMr6lUaNGMDc3V9n/N2zYgA4dOsDCwkLld+Hn54fCwkLs379fZT0DBw6UjmgCT/6eAFTJ3xTw5CiRhYUFZs2aVWabitQ8evRo+Pv7Y8KECRg6dCjc3Nyq9ErFd955B3v37kVqaip2796N1NTUMk+Xbdu2DW3atEH79u2laSYmJhg9ejSuXr2Kc+fOqbQfMWKEyti5yvR1VSyjNuMpM5lp06bNcwdV//PPP7C3ty9xKLVRo0aVXufly5fh4OCgcnqlNAcPHkRYWBgSEhLw8OFDlXmZmZlQKpUVWm/xOJzSam/cuDH+/PNPlWnFY5ueZmFhoTJOpbKSk5MBAF27di11vpmZmfTzgAEDEBISgnXr1uHTTz+FEAIbNmxAz549pXbFy3veqczMzEyVN6Py0NLSQr169VSmNWzYEABUxkkMGzYMwcHB+Oeff+Ds7IwNGzYgPz8fQ4cOrdD6SpOdnQ3gyTgR4Mm2ZmZmwsbGptT2T4/5KEtV71tVxcfHB59//jkUCgWMjIzg7u6uchr33r17mDVrFtauXVtiO58Oco0bN0br1q0RExODkSNHAnhyuuz1118vcQVpafu5UqnEa6+9ViLMKpVKlf0/OTkZp06dKvMU0bM1Ojk5qTwv3h+r4m+quL7JkycjLCwMJ0+eLHV/r2jN3333Hdzc3JCcnIxDhw6phNKX9eabb8LU1BTr1q1DUlISWrdujfr165c6Bumff/6Bj49Pienu7u7S/Kdv9VEVfV3dvy9Nx0BElaZQKFQGvharzCDky5cvo1u3bmjcuDEWLVoER0dH6OnpYdu2bVi8eHGlj3ZUhLa2drUtu7j+H374AXZ2diXm6+j8/5+ig4MDOnTogPXr1+PTTz/F4cOHce3aNemowdPLCw8PL/Pmg9V5k8VBgwZhypQpiImJwaeffooff/wR3t7eLxWci505cwbA/98KoqioCDY2NipHpJ72ovEbVbFvVeXA+qfVqVMHfn5+Zc5/++23cejQIUydOhUtWrSAiYkJioqK0KNHjxJ1Dxs2DJMmTcK///6L3NxcHD58GMuXLy+xzLL287Kmi2cGt7/xxhv46KOPSm1bHJ4rssyXVTyWaNasWaXeY6eiNe/duxe5ubkAngwyf3r81cvS19dHv379sGbNGly5cqXCtxp4nqro65r4fWkyBiJS4ezsjPj4eGRnZ6u8oV64cKFEWwsLi1IPpT57hZSbmxt27NhRYhDu07Zs2YLc3Fxs3rxZ5VNKaadDyntKxtnZWar92SMzFy5ckObXBDc3NwCAjY3Nc98Aiw0cOBDvv/8+Lly4gHXr1sHIyEjlRnTFyzMzMyvX8sqrqKgIV65cUXmTuHjxIgCoXJFiaWmJgIAAxMTEIDAwEAcPHqySG75lZ2dj48aNcHR0lD4Ju7m5YdeuXWjXrl2lPq1XZN+ysLAocUVlXl4ebt269cL1vOypwmfdv38f8fHxmDVrFkJDQ6XpxUcHnzVo0CCEhITgp59+wqNHj6Crq4uBAwdWaU1ubm7Izs6u0n3uZfut+CjRzJkzSz1iWpGab926hQkTJqB79+7Q09PDhx9+CH9/f5X/FS9b7zvvvINVq1ZBS0sLgwYNKrOds7Nzqf93//77b2l+RVX1Pvqq4RgiUvHmm2+ioKBA5ZLkwsJCLFu2rERbNzc3/P3337h9+7Y07a+//ipx2Xf//v0hhCj1PH/xJ4/iTyZPfxLJzMxEdHR0idcYGxuX6y6x3t7esLGxwYoVK6RPfMCTS9/Pnz+PgICAFy6jqvj7+8PMzAxz585Ffn5+iflP9yHwpM+0tbXx008/YcOGDXjrrbdgbGwszffy8oKbmxsWLlwonWJ63vIq4umjCkIILF++HLq6uujWrZtKu6FDh+LcuXOYOnUqtLW1n/vPvTwePXqEoUOH4t69e/jss8+kf95vv/02CgsLMWfOnBKvKSgoeOG+UJF9y83NrcQ4mKioqHIdITI2Nq7S8Uil1Q2gzOBZp04d9OzZEz/++CNiYmLQo0cP6SqoqvL2228jISEBO3bsKDEvIyMDBQUFFV5m8X5d3js/l2by5MkwNzfH7NmzS8yrSM2jRo1CUVERvvvuO0RFRUFHRwcjR45U+R28bL1dunTBnDlzsHz58lKPFhd78803cfToUemWCQCQk5ODqKgouLi4wMPDo8Lrrup99FXDI0Qys337dukTxtPatm2LevXqoVevXmjXrh0++eQTXL16FR4eHvj1119L/SN69913sWjRIvj7+2PkyJFIT0/HihUr0KRJE2RlZUntunTpgqFDh2Lp0qVITk6WDvcfOHAAXbp0QXBwsPSJrFevXhgzZgyys7OxcuVK2NjYlPh07uXlhcjISHz++eeoX78+bGxsSh2bo6uri/nz52PEiBHo1KkTBg8eLF127+LigilTplRBj/6/27dv4/PPPy8x3dXVFYGBgYiMjMTQoUPRqlUrDBo0CNbW1rh27Rp+//13tGvXTiWI2NjYoEuXLli0aBEePHhQ4pO+lpYW/ve//6Fnz55o0qQJRowYgbp16+LGjRvYs2cPzMzMsGXLlgpvg4GBAeLi4hAUFAQfHx9s374dv//+Oz799NMSp6YCAgJgZWUljW8qa4xPaW7cuIEff/wRwJOjQufOncOGDRuQmpqKDz74QOVeRp06dcKYMWMwb948JCUloXv37tDV1UVycjI2bNiAr7/+Gv/973/LXFdF9q333nsPY8eORf/+/fHGG2/gr7/+wo4dO8oVLLy8vLBu3TqEhISgdevWMDExKfXrJcrLzMxMujQ+Pz8fdevWxR9//CHdo6g0w4YNk/qitAD5sqZOnYrNmzfjrbfewvDhw+Hl5YWcnBycPn0aP//8M65evVrhEFZ8p+mJEyfC39+/UuFaqVRi0qRJpX7oKm/N0dHR+P3337F69Wq89tprAIBly5ZhyJAhiIyMxPvvv69S72effYZBgwZBV1cXvXr1UvnA8jxaWlqYPn36C9t98skn+Omnn9CzZ09MnDgRlpaWWLNmDVJSUvDLL79AS6vixzOqeh995ajl2jaqcc+77B7PXE5/9+5dMXToUGFmZiaUSqUYOnSoOHnyZIl2Qgjx448/inr16gk9PT3RokULsWPHjhKX3Qvx5NL98PBw0bhxY6Gnpyesra1Fz549pctlhRBi8+bNolmzZsLAwEC4uLiI+fPni1WrVpW4xDU1NVUEBAQIU1NTAUC6BL+0S6GFEGLdunWiZcuWQl9fX1haWorAwEDx77//qrQp63LtsLCwcl0S3KlTpzL7tvgy2+Ia/f39hVKpFAYGBsLNzU0MHz5cHD9+vMQyV65cKQAIU1NTldsGPO3kyZOiX79+wsrKSujr6wtnZ2fx9ttvi/j4eKlNRS67NzY2FpcvXxbdu3cXRkZGwtbWVoSFhUmX+D7r/fffFwBEbGzsC/uomLOzs9Q3CoVCmJmZiSZNmohRo0aJI0eOlPm6qKgo4eXlJQwNDYWpqanw9PQUH330kbh582aJbXhWefetwsJC8fHHH4s6deoIIyMj4e/vLy5dulSuy+6zs7PFO++8I8zNzQWAF16C/6LbDwghxL///iv69u0rzM3NhVKpFAMGDBA3b94s8/YAubm5wsLCQiiVylL3mbL6p6zbRpRW44MHD8S0adNE/fr1hZ6enqhTp45o27atWLhwocjLyxNClH75e7Fnay8oKBATJkwQ1tbWQqFQvPDvraxa79+/L5RKZanrfVHN169fF0qlUvTq1avEcvv27SuMjY3FlStXpGlz5swRdevWFVpaWi/82yqrz59WVn9dvnxZ/Pe//xXm5ubCwMBAtGnTRmzdulWlTfG+uGHDhlKX+fT/7LL20Yos41WmEEImo6WIqMpNmTIF3333HVJTU2FkZKTucmSvoKAADg4O6NWrF7777jt1l0NUq3AMERFVyuPHj/Hjjz+if//+DEMaYtOmTbh9+zaGDRum7lKIah2OISKiCklPT8euXbvw888/4+7du5g0aZK6S5K9I0eO4NSpU5gzZw5atmyJTp06qbskolqHgYiIKuTcuXMIDAyEjY0Nli5dWuZ9kKjmREZG4scff0SLFi2kL7QloorhGCIiIiKSPY4hIiIiItljICIiIiLZ4xiicigqKsLNmzdhamrKW58TERHVEkIIPHjwAA4ODi+8mSUDUTncvHkTjo6O6i6DiIiIKuH69evSHcjLwkBUDqampgCedKiZmZmaqyEiIqLyyMrKgqOjo/Q+/jwMROVQfJrMzMyMgYiIiKiWKc9wFw6qJiIiItljICIiIiLZYyAiIiIi2eMYIiKialZYWIj8/Hx1l0H0StLT03vhJfXlwUBERFRNhBBITU1FRkaGukshemVpaWnB1dUVenp6L7UcBiIiompSHIZsbGxgZGTEG7sSVbHiGyffunULTk5OL/U3xkBERFQNCgsLpTBkZWWl7nKIXlnW1ta4efMmCgoKoKurW+nlcFA1EVE1KB4zZGRkpOZKiF5txafKCgsLX2o5DERERNWIp8mIqldV/Y0xEBEREZHsMRAREWmQzp07Y/LkyeVuv2nTJtSvXx/a2toVet2LKBQKbNq0qcqWR1Rew4cPR58+fWp8vQxERES12JgxY/Df//4X169fx5w5c6plHVevXoVCoUBSUlK1LL820tQ+Wb16NczNzdVdRrloWh/yKjMioloqOzsb6enp8Pf3h4ODg7rLIarVeISIiEhNcnJyMGzYMJiYmMDe3h5fffWVyvzc3Fx8+OGHqFu3LoyNjeHj44O9e/cCAPbu3QtTU1MAQNeuXaFQKLB3717cvXsXgwcPRt26dWFkZARPT0/89NNPKst1cXHBkiVLVKa1aNECM2fOLLVOV1dXAEDLli2hUCjQuXPnF25b8WmPWbNmwdraGmZmZhg7dizy8vKkNnFxcWjfvj3Mzc1hZWWFt956C5cvX5bmd+3aFcHBwSrLvX37NvT09BAfHy9ty+effy71o7OzMzZv3ozbt2+jd+/eMDExQbNmzXD8+HGV5fz555/o0KEDDA0N4ejoiIkTJyInJ0elj+bOnYt3330XpqamcHJyQlRUVJX0ycKFC2Fvbw8rKyuMHz9e5S7mz/udP378GE2aNMHo0aOl9pcvX4apqSlWrVqFvXv3YsSIEcjMzIRCoYBCoSjzd/q0yvbhL7/8giZNmkBfXx8uLi4l9t+X7cPn9VO1EPRCmZmZAoDIzMxUdylEVEs8evRInDt3Tjx69KjMNuPGjRNOTk5i165d4tSpU+Ktt94SpqamYtKkSUIIId577z3Rtm1bsX//fnHp0iURHh4u9PX1xcWLF0Vubq64cOGCACB++eUXcevWLZGbmyv+/fdfER4eLk6ePCkuX74sli5dKrS1tcWRI0ek9To7O4vFixer1NK8eXMRFhYmPQcgNm7cKIQQ4ujRowKA2LVrl7h165a4e/fuC7c/KChImJiYiIEDB4ozZ86IrVu3Cmtra/Hpp59KbX7++Wfxyy+/iOTkZHHy5EnRq1cv4enpKQoLC4UQQsTExAgLCwvx+PFj6TWLFi0SLi4uoqioSNoWS0tLsWLFCnHx4kUxbtw4YWZmJnr06CHWr18vLly4IPr06SPc3d2l11y6dEkYGxuLxYsXi4sXL4qDBw+Kli1biuHDh6v0kaWlpYiIiBDJycli3rx5QktLS/z9998v1SdmZmZi7Nix4vz582LLli3CyMhIREVFSW2e9zsXQoiTJ08KPT09sWnTJlFQUCBef/110bdvXyGEELm5uWLJkiXCzMxM3Lp1S9y6dUs8ePDghXVVpg+PHz8utLS0xOzZs8WFCxdEdHS0MDQ0FNHR0S/dh+Xpp6c972+tIu/fDETlwEBE1aHVh2tq/EE150WB6MGDB0JPT0+sX79emnb37l1haGgoJk2aJP755x+hra0tbty4ofK6bt26iWnTpgkhhLh//74AIPbs2fPcWgICAsQHH3wgPa9oIEpJSREAxMmTJ5+/0U8JCgoSlpaWIicnR5oWGRkpTExMpMDzrNu3bwsA4vTp00KIJ31oYWEh1q1bJ7Vp1qyZmDlzpsq2DBkyRHp+69YtAUDMmDFDmpaQkCAAiFu3bgkhhBg5cqQYPXq0yroPHDggtLS0pN/Xs8stKioSNjY2IjIy8qX6xNnZWRQUFEjTBgwYIAYOHCiEEOX6nQshxIIFC0SdOnVEcHCwsLe3F3fu3JHmRUdHC6VSWe6ahKhcH77zzjvijTfeUFnO1KlThYeHR5nLLW8fvqifnlVVgYinzIiI1ODy5cvIy8uDj4+PNM3S0hKNGjUCAJw+fRqFhYVo2LAhTExMpMe+fftUTis9q7CwEHPmzIGnpycsLS1hYmKCHTt24Nq1a9W+Tc9q3ry5yo0pfX19kZ2djevXrwMAkpOTMXjwYNSrVw9mZmZwcXEBAKlWAwMDDB06FKtWrQIAnDhxAmfOnMHw4cNV1tOsWTPpZ1tbWwCAp6dniWnp6ekAgL/++gurV69W6Vd/f38UFRUhJSWl1OUqFArY2dlJy6isJk2aQFtbW3pub28vLbO8v/MPPvgADRs2xPLly7Fq1aoquRN6Rfvw/PnzaNeuncoy2rVrh+TkZJUbJFa2D5/XT9WFg6qJiDRQdnY2tLW1kZiYqPLGAAAmJiZlvi48PBxff/01lixZAk9PTxgbG2Py5MkqY3e0tLQghFB5XbWPzyhFr1694OzsjJUrV8LBwQFFRUVo2rSpSq3vvfceWrRogX///RfR0dHo2rUrnJ2dVZbz9Nc1FN+kr7RpRUVFAJ707ZgxYzBx4sQSNTk5OZW63OLlFC+jsp63zPL+ztPT03Hx4kVoa2sjOTkZPXr0eKmanq2rPH1YmeUWL6c8y6iOvn8RBiIiIjVwc3ODrq4ujhw5Ir0J379/HxcvXkSnTp3QsmVLFBYWIj09HR06dCj3cg8ePIjevXtjyJAhAJ68gV28eBEeHh5SG2tra9y6dUt6npWVpXJk5FmV/WqEv/76C48ePYKhoSEA4PDhwzAxMYGjoyPu3r2LCxcuYOXKldL2/fnnnyWW4enpCW9vb6xcuRKxsbFYvnx5hWooTatWrXDu3DnUr1+/0suoqq+LeFp5f+fvvvsuPD09MXLkSIwaNQp+fn5wd3eX6qrKmsri7u6OgwcPqkw7ePAgGjZsWCLMlaU6+vBl8JQZEZEamJiYYOTIkZg6dSp2794tnQrS0nryb7lhw4YIDAzEsGHD8OuvvyIlJQVHjx7FvHnz8Pvvv5e53AYNGmDnzp04dOgQzp8/jzFjxiAtLU2lTdeuXfHDDz/gwIEDOH36NIKCgp77JmZjYwNDQ0PExcUhLS0NmZmZ5drGvLw8jBw5EufOncO2bdsQFhaG4OBgaGlpwcLCAlZWVoiKisKlS5ewe/duhISElLqc9957D19++SWEEOjbt2+51v08H3/8MQ4dOoTg4GAkJSUhOTkZv/32W4kr2p6nsn3yPOX5nUdERCAhIQFr1qxBYGAg+vTpg8DAQOmomouLC7KzsxEfH487d+7g4cOHL11XaT744APEx8djzpw5uHjxItasWYPly5fjww8/LPcyqqMPXwYDERGRmoSHh6NDhw7o1asX/Pz80L59e3h5eUnzo6OjMWzYMHzwwQdo1KgR+vTpg2PHjqmc1nnW9OnT0apVK/j7+6Nz586ws7MrcdffadOmoVOnTnjrrbcQEBCAPn36wM3Nrcxl6ujoYOnSpfj222/h4OCA3r17l2v7unXrhgYNGqBjx44YOHAg/vOf/0iXgWtpaWHt2rVITExE06ZNMWXKFISHh5e6nMGDB0NHRweDBw+GgYFBudb9PM2aNcO+fftw8eJFdOjQAS1btkRoaGiF7uVU2T55kef9zv/++29MnToV33zzDRwdHQEA33zzDe7cuYMZM2YAANq2bYuxY8di4MCBsLa2xoIFC6qkrme1atUK69evx9q1a9G0aVOEhoZi9uzZJcZ3PU919WFlKcSzJ5KphKysLCiVSmRmZsLMzEzd5dArwmvq9zW+zsTwYTW+Trl6/PgxUlJS4OrqWiVv4rXN8OHDkZGRUSVf/3H16lW4ubnh2LFjaNWq1csXR6+U5/2tVeT9m2OIiIhII+Xn5+Pu3buYPn06Xn/9dYYhqlYMREREVGHPu9Jt+/btVbKOgwcPokuXLmjYsCF+/vnnKllmdXpRn1RkcHxVOXDgAHr27Fnm/Ozs7BqsRrMxEBERUYU97ws569atWyVv/p07dy5xewBN9qI+UQdvb2+N+fJUTcdAREREFfYyl6y/qjSxTwwNDTWyLk3Eq8yIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9tQaiPbv349evXrBwcEBCoWixA28hBAIDQ2Fvb09DA0N4efnh+TkZJU29+7dQ2BgIMzMzGBubo6RI0eWuIzw1KlT6NChAwwMDODo6Fhtd+4kIiKi2kmtgSgnJwfNmzdHREREqfMXLFiApUuXYsWKFThy5AiMjY3h7++Px48fS20CAwNx9uxZ7Ny5E1u3bsX+/fsxevRoaX5WVha6d+8OZ2dnJCYmIjw8HDNnzkRUVFS1bx8RERHVDmq97L5nz55l3jBKCIElS5Zg+vTp0vebfP/997C1tcWmTZswaNAgnD9/HnFxcTh27Bi8vb0BAMuWLcObb76JhQsXwsHBATExMcjLy8OqVaugp6eHJk2aICkpCYsWLVIJTkREVDvU9NfeVPYrbyIiIhAeHo7U1FQ0b94cy5YtQ5s2baq4OqoqGjuGKCUlBampqfDz85OmKZVK+Pj4ICEhAQCQkJAAc3NzKQwBgJ+fH7S0tHDkyBGpTceOHaGnpye18ff3x4ULF3D//v1S152bm4usrCyVBxERUXmtW7cOISEhCAsLw4kTJ9C8eXP4+/sjPT1d3aVRGTQ2EKWmpgIAbG1tVabb2tpK81JTU2FjY6MyX0dHB5aWliptSlvG0+t41rx586BUKqVH8bcKExERlceiRYswatQojBgxAh4eHlixYgWMjIywatUqdZdGZdDYQKRO06ZNQ2ZmpvS4fv26uksiIqJaIi8vD4mJiSpnOLS0tODn5yed4SDNo7GByM7ODgCQlpamMj0tLU2aZ2dnV+LwY0FBAe7du6fSprRlPL2OZ+nr68PMzEzlQUREVB537txBYWHhc89wkObR2EDk6uoKOzs7xMfHS9OysrJw5MgR+Pr6AgB8fX2RkZGBxMREqc3u3btRVFQEHx8fqc3+/fuRn58vtdm5cycaNWoECwuLGtoaIiIi0mRqDUTZ2dlISkqSvok3JSUFSUlJuHbtGhQKBSZPnozPP/8cmzdvxunTpzFs2DA4ODigT58+AAB3d3f06NEDo0aNwtGjR3Hw4EEEBwdj0KBBcHBwAAC888470NPTw8iRI3H27FmsW7cOX3/9NUJCQtS01URE9CqrU6cOtLW1n3uGgzSPWgPR8ePH0bJlS7Rs2RIAEBISgpYtWyI0NBQA8NFHH2HChAkYPXo0WrdujezsbMTFxcHAwEBaRkxMDBo3boxu3brhzTffRPv27VXuMaRUKvHHH38gJSUFXl5e+OCDDxAaGspL7omIqFro6enBy8tL5QxHUVER4uPjpTMcpHnUeh+izp07QwhR5nyFQoHZs2dj9uzZZbaxtLREbGzsc9fTrFkzHDhwoNJ1EhERVURISAiCgoLg7e2NNm3aYMmSJcjJycGIESPUXRqVQa2BiIiI6FU0cOBA3L59G6GhoUhNTUWLFi0QFxdXYqA1aQ4GIiIiqlUqe+fomhYcHIzg4GB1l0HlpLFXmRERERHVFAYiIiIikj0GIiIiIpI9BiIiIiKSPQYiIiIikj0GIiIiIpI9BiIiIiKSPQYiIiIikj0GIiIiIpI9BiIiIiKSPX51BxER1SrXZnvW6PqcQk9X+DX79+9HeHg4EhMTcevWLWzcuBF9+vSp+uKoyvAIERERURXLyclB8+bNERERoe5SqJx4hIiIiKiK9ezZEz179lR3GVQBPEJEREREssdARERERLLHQERERESyx0BEREREssdARERERLLHq8yIiIiqWHZ2Ni5duiQ9T0lJQVJSEiwtLeHk5KTGyqgsDERERERV7Pjx4+jSpYv0PCQkBAAQFBSE1atXq6kqeh4GIiIiqlUqc+fomta5c2cIIdRdBlUAxxARERGR7DEQERERkewxEBEREZHsMRARERGR7DEQERFVIw6sJapeVfU3xkBERFQNdHV1AQAPHz5UcyVEr7a8vDwAgLa29ksth5fdExFVA21tbZibmyM9PR0AYGRkBIVCoeaqiF4tRUVFuH37NoyMjKCj83KRhoGIiKia2NnZAYAUioio6mlpacHJyemlP3AwEBERVROFQgF7e3vY2NggPz9f3eUQvZL09PSgpfXyI4AYiIiIqpm2tvZLj28gourFQdVEREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7vOyeqpXX1O9rfJ2J4cNqfJ1ERFS78QgRERERyR4DEREREckeAxERERHJHgMRERERyR4DEREREckeAxERERHJHgMRERERyR4DEREREckeAxERERHJHgMRERERyR4DEREREckeAxERERHJHgMRERERyR4DEREREckeAxERERHJnkYHosLCQsyYMQOurq4wNDSEm5sb5syZAyGE1EYIgdDQUNjb28PQ0BB+fn5ITk5WWc69e/cQGBgIMzMzmJubY+TIkcjOzq7pzSEiIiINpdGBaP78+YiMjMTy5ctx/vx5zJ8/HwsWLMCyZcukNgsWLMDSpUuxYsUKHDlyBMbGxvD398fjx4+lNoGBgTh79ix27tyJrVu3Yv/+/Rg9erQ6NomIiIg0kI66C3ieQ4cOoXfv3ggICAAAuLi44KeffsLRo0cBPDk6tGTJEkyfPh29e/cGAHz//fewtbXFpk2bMGjQIJw/fx5xcXE4duwYvL29AQDLli3Dm2++iYULF8LBwUE9G0dEREQaQ6OPELVt2xbx8fG4ePEiAOCvv/7Cn3/+iZ49ewIAUlJSkJqaCj8/P+k1SqUSPj4+SEhIAAAkJCTA3NxcCkMA4OfnBy0tLRw5cqQGt4aIiIg0lUYfIfrkk0+QlZWFxo0bQ1tbG4WFhfjiiy8QGBgIAEhNTQUA2NraqrzO1tZWmpeamgobGxuV+To6OrC0tJTaPCs3Nxe5ubnS86ysrCrbJiIiItI8Gn2EaP369YiJiUFsbCxOnDiBNWvWYOHChVizZk21rnfevHlQKpXSw9HRsVrXR0REROql0YFo6tSp+OSTTzBo0CB4enpi6NChmDJlCubNmwcAsLOzAwCkpaWpvC4tLU2aZ2dnh/T0dJX5BQUFuHfvntTmWdOmTUNmZqb0uH79elVvGhEREWkQjQ5EDx8+hJaWaona2tooKioCALi6usLOzg7x8fHS/KysLBw5cgS+vr4AAF9fX2RkZCAxMVFqs3v3bhQVFcHHx6fU9err68PMzEzlQURERK8ujR5D1KtXL3zxxRdwcnJCkyZNcPLkSSxatAjvvvsuAEChUGDy5Mn4/PPP0aBBA7i6umLGjBlwcHBAnz59AADu7u7o0aMHRo0ahRUrViA/Px/BwcEYNGgQrzAjIiIiABoeiJYtW4YZM2bg/fffR3p6OhwcHDBmzBiEhoZKbT766CPk5ORg9OjRyMjIQPv27REXFwcDAwOpTUxMDIKDg9GtWzdoaWmhf//+WLp0qTo2iYiIiDSQQjx922cqVVZWFpRKJTIzM3n6rIK8pn5f4+tMDB9W4+usDPYNEVH1qsj7t0aPISIiIiKqCQxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsMRERERCR7DEREREQkewxEREREJHsaH4hu3LiBIUOGwMrKCoaGhvD09MTx48el+UIIhIaGwt7eHoaGhvDz80NycrLKMu7du4fAwECYmZnB3NwcI0eORHZ2dk1vChEREWmoSgWievXq4e7duyWmZ2RkoF69ei9dVLH79++jXbt20NXVxfbt23Hu3Dl89dVXsLCwkNosWLAAS5cuxYoVK3DkyBEYGxvD398fjx8/ltoEBgbi7Nmz2LlzJ7Zu3Yr9+/dj9OjRVVYnERER1W46lXnR1atXUVhYWGJ6bm4ubty48dJFFZs/fz4cHR0RHR0tTXN1dZV+FkJgyZIlmD59Onr37g0A+P7772Fra4tNmzZh0KBBOH/+POLi4nDs2DF4e3sDAJYtW4Y333wTCxcuhIODQ5XVS0RERLVThQLR5s2bpZ937NgBpVIpPS8sLER8fDxcXFyqrLjNmzfD398fAwYMwL59+1C3bl28//77GDVqFAAgJSUFqamp8PPzk16jVCrh4+ODhIQEDBo0CAkJCTA3N5fCEAD4+flBS0sLR44cQd++fUusNzc3F7m5udLzrKysKtsmIiIi0jwVCkR9+vQBACgUCgQFBanM09XVhYuLC7766qsqK+7KlSuIjIxESEgIPv30Uxw7dgwTJ06Enp4egoKCkJqaCgCwtbVVeZ2tra00LzU1FTY2NirzdXR0YGlpKbV51rx58zBr1qwq2w4iIiLSbBUKREVFRQCenLY6duwY6tSpUy1FPb0+b29vzJ07FwDQsmVLnDlzBitWrCgRyKrStGnTEBISIj3PysqCo6Njta2PiIiI1KtSg6pTUlKqPQwBgL29PTw8PFSmubu749q1awAAOzs7AEBaWppKm7S0NGmenZ0d0tPTVeYXFBTg3r17Uptn6evrw8zMTOVBREREr65KDaoGgPj4eMTHxyM9PV06clRs1apVL10YALRr1w4XLlxQmXbx4kU4OzsDeHKkys7ODvHx8WjRogWAJ0dzjhw5gnHjxgEAfH19kZGRgcTERHh5eQEAdu/ejaKiIvj4+FRJnURERFS7VSoQzZo1C7Nnz4a3tzfs7e2hUCiqui4AwJQpU9C2bVvMnTsXb7/9No4ePYqoqChERUUBeDKWafLkyfj888/RoEEDuLq6YsaMGXBwcJDGO7m7u6NHjx4YNWoUVqxYgfz8fAQHB2PQoEG8woyIiIgAVDIQrVixAqtXr8bQoUOruh4VrVu3xsaNGzFt2jTMnj0brq6uWLJkCQIDA6U2H330EXJycjB69GhkZGSgffv2iIuLg4GBgdQmJiYGwcHB6NatG7S0tNC/f38sXbq0WmsnIiKi2kMhhBAVfZGVlRWOHj0KNze36qhJ42RlZUGpVCIzM5PjiSrIa+r3Nb7OxPBhNb7OymDfEBFVr4q8f1dqUPV7772H2NjYShVHREREpGkqdcrs8ePHiIqKwq5du9CsWTPo6uqqzF+0aFGVFEdERERUEyoViE6dOiVd1XXmzBmVedU1wJqIiIioulQqEO3Zs6eq6yAiIiJSm0qNISIiIiJ6lVTqCFGXLl2ee2ps9+7dlS6IiIiIqKZVKhAVjx8qlp+fj6SkJJw5c6Zav2OMiIiIqDpUKhAtXry41OkzZ85Ednb2SxVEREREVNOqdAzRkCFDqux7zIiIiIhqSpUGooSEBJWvzCAiIiKqDSp1yqxfv34qz4UQuHXrFo4fP44ZM2ZUSWFERERENaVSgUipVKo819LSQqNGjTB79mx07969SgojIiIiqimVCkTR0dFVXQcRERGR2lQqEBVLTEzE+fPnAQBNmjRBy5Ytq6QoIiIioppUqUCUnp6OQYMGYe/evTA3NwcAZGRkoEuXLli7di2sra2rskYiIiKialWpq8wmTJiABw8e4OzZs7h37x7u3buHM2fOICsrCxMnTqzqGomIiIiqVaWOEMXFxWHXrl1wd3eXpnl4eCAiIoKDqomIiKjWqdQRoqKiIujq6paYrquri6KiopcuioiIiKgmVSoQde3aFZMmTcLNmzelaTdu3MCUKVPQrVu3KiuOiIiIqCZUKhAtX74cWVlZcHFxgZubG9zc3ODq6oqsrCwsW7asqmskIiIiqlaVGkPk6OiIEydOYNeuXfj7778BAO7u7vDz86vS4oiIiIhqQoWOEO3evRseHh7IysqCQqHAG2+8gQkTJmDChAlo3bo1mjRpggMHDlRXrURERETVokKBaMmSJRg1ahTMzMxKzFMqlRgzZgwWLVpUZcURERER1YQKBaK//voLPXr0KHN+9+7dkZiY+NJFEREREdWkCgWitLS0Ui+3L6ajo4Pbt2+/dFFERERENalCgahu3bo4c+ZMmfNPnToFe3v7ly6KiIiIqCZVKBC9+eabmDFjBh4/flxi3qNHjxAWFoa33nqryoojIiIiqgkVuux++vTp+PXXX9GwYUMEBwejUaNGAIC///4bERERKCwsxGeffVYthRIRERFVlwoFIltbWxw6dAjjxo3DtGnTIIQAACgUCvj7+yMiIgK2trbVUigRERFRdanwjRmdnZ2xbds23L9/H5cuXYIQAg0aNICFhUV11EdERERU7Sp1p2oAsLCwQOvWrauyFiIiIiK1qNR3mRERERG9ShiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2alUg+vLLL6FQKDB58mRp2uPHjzF+/HhYWVnBxMQE/fv3R1pamsrrrl27hoCAABgZGcHGxgZTp05FQUFBDVdPREREmqrWBKJjx47h22+/RbNmzVSmT5kyBVu2bMGGDRuwb98+3Lx5E/369ZPmFxYWIiAgAHl5eTh06BDWrFmD1atXIzQ0tKY3gYiIiDRUrQhE2dnZCAwMxMqVK2FhYSFNz8zMxHfffYdFixaha9eu8PLyQnR0NA4dOoTDhw8DAP744w+cO3cOP/74I1q0aIGePXtizpw5iIiIQF5enro2iYiIiDRIrQhE48ePR0BAAPz8/FSmJyYmIj8/X2V648aN4eTkhISEBABAQkICPD09YWtrK7Xx9/dHVlYWzp49WzMbQERERBpNR90FvMjatWtx4sQJHDt2rMS81NRU6OnpwdzcXGW6ra0tUlNTpTZPh6Hi+cXzSpObm4vc3FzpeVZW1stsAhEREWk4jT5CdP36dUyaNAkxMTEwMDCosfXOmzcPSqVSejg6OtbYuomIiKjmaXQgSkxMRHp6Olq1agUdHR3o6Ohg3759WLp0KXR0dGBra4u8vDxkZGSovC4tLQ12dnYAADs7uxJXnRU/L27zrGnTpiEzM1N6XL9+veo3joiIiDSGRgeibt264fTp00hKSpIe3t7eCAwMlH7W1dVFfHy89JoLFy7g2rVr8PX1BQD4+vri9OnTSE9Pl9rs3LkTZmZm8PDwKHW9+vr6MDMzU3kQERHRq0ujxxCZmpqiadOmKtOMjY1hZWUlTR85ciRCQkJgaWkJMzMzTJgwAb6+vnj99dcBAN27d4eHhweGDh2KBQsWIDU1FdOnT8f48eOhr69f49tEREREmkejA1F5LF68GFpaWujfvz9yc3Ph7++Pb775Rpqvra2NrVu3Yty4cfD19YWxsTGCgoIwe/ZsNVZNREREmqTWBaK9e/eqPDcwMEBERAQiIiLKfI2zszO2bdtWzZURERFRbaXRY4iIiIiIagIDEREREckeAxERERHJHgMRERERyV6tG1RNRK8+r6nf1+j6EsOH1ej6iEjz8AgRERERyR4DEREREckeAxERERHJHgMRERERyR4DEREREckeAxERERHJHgMRERERyR4DEREREckeb8xIJCPXZnvW6PqcQk/X6PqIiCqLR4iIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2GIiIiIhI9hiIiIiISPYYiIiIiEj2NDoQzZs3D61bt4apqSlsbGzQp08fXLhwQaXN48ePMX78eFhZWcHExAT9+/dHWlqaSptr164hICAARkZGsLGxwdSpU1FQUFCTm0JEREQaTKMD0b59+zB+/HgcPnwYO3fuRH5+Prp3746cnBypzZQpU7BlyxZs2LAB+/btw82bN9GvXz9pfmFhIQICApCXl4dDhw5hzZo1WL16NUJDQ9WxSURERKSBdNRdwPPExcWpPF+9ejVsbGyQmJiIjh07IjMzE9999x1iY2PRtWtXAEB0dDTc3d1x+PBhvP766/jjjz9w7tw57Nq1C7a2tmjRogXmzJmDjz/+GDNnzoSenp46No2IiIg0iEYfIXpWZmYmAMDS0hIAkJiYiPz8fPj5+UltGjduDCcnJyQkJAAAEhIS4OnpCVtbW6mNv78/srKycPbs2VLXk5ubi6ysLJUHERERvbpqTSAqKirC5MmT0a5dOzRt2hQAkJqaCj09PZibm6u0tbW1RWpqqtTm6TBUPL94XmnmzZsHpVIpPRwdHat4a4iIiEiT1JpANH78eJw5cwZr166t9nVNmzYNmZmZ0uP69evVvk4iIiJSH40eQ1QsODgYW7duxf79+/Haa69J0+3s7JCXl4eMjAyVo0RpaWmws7OT2hw9elRlecVXoRW3eZa+vj709fWreCuIiIhIU2n0ESIhBIKDg7Fx40bs3r0brq6uKvO9vLygq6uL+Ph4adqFCxdw7do1+Pr6AgB8fX1x+vRppKenS2127twJMzMzeHh41MyGEBERkUbT6CNE48ePR2xsLH777TeYmppKY36USiUMDQ2hVCoxcuRIhISEwNLSEmZmZpgwYQJ8fX3x+uuvAwC6d+8ODw8PDB06FAsWLEBqaiqmT5+O8ePH8ygQERERAdDwQBQZGQkA6Ny5s8r06OhoDB8+HACwePFiaGlpoX///sjNzYW/vz+++eYbqa22tja2bt2KcePGwdfXF8bGxggKCsLs2bNrajOIiIhIw2l0IBJCvLCNgYEBIiIiEBERUWYbZ2dnbNu2rSpLIyIioleIRo8hIiIiIqoJGn2EiKgyrs32rNH1OYWertH1ERFR1eMRIiIiIpI9BiIiIiKSPQYiIiIikj0GIiIiIpI9BiIiIiKSPQYiIiIikj0GIiIiIpI9BiIiIiKSPQYiIiIikj0GIiIiIpI9BiIiIiKSPQYiIiIikj0GIiIiIpI9BiIiIiKSPQYiIiIikj0GIiIiIpI9BiIiIiKSPQYiIiIikj0GIiIiIpI9BiIiIiKSPR11F0BEpG7XZnvW+DqdQk/X+DqJqGw8QkRERESyx0BEREREssdARERERLLHQERERESyx0BEREREssdARERERLLHQERERESyx0BEREREssdARERERLLHQERERESyx0BEREREssdARERERLLHQERERESyx0BEREREssdARERERLLHQERERESyx0BEREREssdARERERLLHQERERESyx0BEREREssdARERERLLHQERERESyx0BEREREssdARERERLLHQERERESyx0BEREREssdARERERLLHQERERESyx0BEREREsqej7gKIiIiqgtfU72t0fYnhw2p0fVS9eISIiIiIZI+BiIiIiGRPVqfMIiIiEB4ejtTUVDRv3hzLli1DmzZt1F0WEVG58bQQUfWQzRGidevWISQkBGFhYThx4gSaN28Of39/pKenq7s0IiIiUjPZBKJFixZh1KhRGDFiBDw8PLBixQoYGRlh1apV6i6NiIiI1EwWgSgvLw+JiYnw8/OTpmlpacHPzw8JCQlqrIyIiIg0gSzGEN25cweFhYWwtbVVmW5ra4u///67RPvc3Fzk5uZKzzMzMwEAWVlZpS6/4/SfqrDa8tn/+eAaX2dlFOY+qvF1PtAtrNH1lbVfvAj7pmw13Tc13S9A7embytapDjXdN2c/9ajR9Tl+crjSr63p96kYk69rdH1l9U3x/iuEePFChAzcuHFDABCHDh1SmT516lTRpk2bEu3DwsIEAD744IMPPvjg4xV4XL9+/YVZQRZHiOrUqQNtbW2kpaWpTE9LS4OdnV2J9tOmTUNISIj0vKioCPfu3YOVlRUUCkW11/siWVlZcHR0xPXr12FmZqbucjQK+6Zs7JvSsV/Kxr4pG/umbJrUN0IIPHjwAA4ODi9sK4tApKenBy8vL8THx6NPnz4AnoSc+Ph4BAcHl2ivr68PfX19lWnm5uY1UGnFmJmZqX1n01Tsm7Kxb0rHfikb+6Zs7JuyaUrfKJXKcrWTRSACgJCQEAQFBcHb2xtt2rTBkiVLkJOTgxEjRqi7NCIiIlIz2QSigQMH4vbt2wgNDUVqaipatGiBuLi4EgOtiYiISH5kE4gAIDg4uNRTZLWNvr4+wsLCSpzWI/bN87BvSsd+KRv7pmzsm7LV1r5RCFGea9GIiIiIXl2yuDEjERER0fMwEBEREZHsMRARERGR7DEQERERkewxENUyERERcHFxgYGBAXx8fHD06FF1l6QR9u/fj169esHBwQEKhQKbNm1Sd0kaYd68eWjdujVMTU1hY2ODPn364MKFC+ouSyNERkaiWbNm0s3jfH19sX37dnWXpZG+/PJLKBQKTJ48Wd2lqN3MmTOhUChUHo0bN1Z3WRrjxo0bGDJkCKysrGBoaAhPT08cP35c3WWVCwNRLbJu3TqEhIQgLCwMJ06cQPPmzeHv74/09HR1l6Z2OTk5aN68OSIiItRdikbZt28fxo8fj8OHD2Pnzp3Iz89H9+7dkZOTo+7S1O61117Dl19+icTERBw/fhxdu3ZF7969cfbsWXWXplGOHTuGb7/9Fs2aNVN3KRqjSZMmuHXrlvT4888/1V2SRrh//z7atWsHXV1dbN++HefOncNXX30FCwsLdZdWLrzsvhbx8fFB69atsXz5cgBPvn7E0dEREyZMwCeffKLm6jSHQqHAxo0bpa9pof93+/Zt2NjYYN++fejYsaO6y9E4lpaWCA8Px8iRI9VdikbIzs5Gq1at8M033+Dzzz9HixYtsGTJEnWXpVYzZ87Epk2bkJSUpO5SNM4nn3yCgwcP4sCBA+oupVJ4hKiWyMvLQ2JiIvz8/KRpWlpa8PPzQ0JCghoro9okMzMTwJM3fvp/hYWFWLt2LXJycuDr66vucjTG+PHjERAQoPJ/h4Dk5GQ4ODigXr16CAwMxLVr19RdkkbYvHkzvL29MWDAANjY2KBly5ZYuXKlussqNwaiWuLOnTsoLCws8VUjtra2SE1NVVNVVJsUFRVh8uTJaNeuHZo2barucjTC6dOnYWJiAn19fYwdOxYbN26Eh4eHusvSCGvXrsWJEycwb948dZeiUXx8fLB69WrExcUhMjISKSkp6NChAx48eKDu0tTuypUriIyMRIMGDbBjxw6MGzcOEydOxJo1a9RdWrnI6qs7iORs/PjxOHPmDMc7PKVRo0ZISkpCZmYmfv75ZwQFBWHfvn2yD0XXr1/HpEmTsHPnThgYGKi7HI3Ss2dP6edmzZrBx8cHzs7OWL9+vexPtRYVFcHb2xtz584FALRs2RJnzpzBihUrEBQUpObqXoxHiGqJOnXqQFtbG2lpaSrT09LSYGdnp6aqqLYIDg7G1q1bsWfPHrz22mvqLkdj6OnpoX79+vDy8sK8efPQvHlzfP311+ouS+0SExORnp6OVq1aQUdHBzo6Oti3bx+WLl0KHR0dFBYWqrtEjWFubo6GDRvi0qVL6i5F7ezt7Ut8mHB3d681pxQZiGoJPT09eHl5IT4+XppWVFSE+Ph4jnmgMgkhEBwcjI0bN2L37t1wdXVVd0karaioCLm5ueouQ+26deuG06dPIykpSXp4e3sjMDAQSUlJ0NbWVneJGiM7OxuXL1+Gvb29uktRu3bt2pW4rcfFixfh7OyspooqhqfMapGQkBAEBQXB29sbbdq0wZIlS5CTk4MRI0aouzS1y87OVvmElpKSgqSkJFhaWsLJyUmNlanX+PHjERsbi99++w2mpqbSeDOlUglDQ0M1V6de06ZNQ8+ePeHk5IQHDx4gNjYWe/fuxY4dO9RdmtqZmpqWGGdmbGwMKysr2Y8/+/DDD9GrVy84Ozvj5s2bCAsLg7a2NgYPHqzu0tRuypQpaNu2LebOnYu3334bR48eRVRUFKKiotRdWvkIqlWWLVsmnJychJ6enmjTpo04fPiwukvSCHv27BEASjyCgoLUXZpaldYnAER0dLS6S1O7d999Vzg7Ows9PT1hbW0tunXrJv744w91l6WxOnXqJCZNmqTuMtRu4MCBwt7eXujp6Ym6deuKgQMHikuXLqm7LI2xZcsW0bRpU6Gvry8aN24soqKi1F1SufE+RERERCR7HENEREREssdARERERLLHQERERESyx0BEREREssdARERERLLHQERERESyx0BEREREssdARETVTqFQYNOmTeouA8OHD0efPn3UXUa1W716NczNzdVdBlGtwkBEROU2fPhwKBSKEo8ePXqouzQVV69ehUKhQFJSksr0r7/+GqtXr6729WtKACSi8uN3mRFRhfTo0QPR0dEq0/T19dVUTcUolUp1l0BEGopHiIioQvT19WFnZ6fysLCwkOYnJyejY8eOMDAwgIeHB3bu3Kny+r1790KhUCAjI0OalpSUBIVCgatXr0rTDh48iM6dO8PIyAgWFhbw9/fH/fv3AQBxcXFo3749zM3NYWVlhbfeeguXL1+WXuvq6goAaNmyJRQKBTp37gyg5Cmz3NxcTJw4ETY2NjAwMED79u1x7NixErXGx8fD29sbRkZGaNu2bYlv9K6o//3vf3B3d4eBgQEaN26Mb775RprXtm1bfPzxxyrtb9++DV1dXezfv1+q+8MPP0TdunVhbGwMHx8f7N2796VqIpI7BiIiqjJFRUXo168f9PT0cOTIEaxYsaLEm3t5JCUloVu3bvDw8EBCQgL+/PNP9OrVC4WFhQCAnJwchISE4Pjx44iPj4eWlhb69u2LoqIiAMDRo0cBALt27cKtW7fw66+/lrqejz76CL/88gvWrFmDEydOoH79+vD398e9e/dU2n322Wf46quvcPz4cejo6ODdd9+t8DYVi4mJQWhoKL744gucP38ec+fOxYwZM7BmzRoAQGBgINauXYunv2Zy3bp1cHBwQIcOHQAAwcHBSEhIwNq1a3Hq1CkMGDAAPXr0QHJycqXrIpI9NX+5LBHVIkFBQUJbW1sYGxurPL744gshhBA7duwQOjo64saNG9Jrtm/fLgCIjRs3CiGE2LNnjwAg7t+/L7U5efKkACBSUlKEEEIMHjxYtGvXrtx13b59WwAQp0+fFkIIkZKSIgCIkydPlqi/d+/eQgghsrOzha6uroiJiZHm5+XlCQcHB7FgwQKVWnft2iW1+f333wUA8ejRozLreXp7n+Xm5iZiY2NVps2ZM0f4+voKIYRIT08XOjo6Yv/+/dJ8X19f8fHHHwshhPjnn3+Etra2Sh8LIUS3bt3EtGnThBBCREdHC6VSWWZ9RFQSxxARUYV06dIFkZGRKtMsLS0BAOfPn4ejoyMcHBykeb6+vhVeR1JSEgYMGFDm/OTkZISGhuLIkSO4c+eOdGTo2rVraNq0abnWcfnyZeTn56Ndu3bSNF1dXbRp0wbnz59XadusWTPpZ3t7ewBAeno6nJycyr1NwJMjW5cvX8bIkSMxatQoaXpBQYE0vsna2hrdu3dHTEwMOnTogJSUFCQkJODbb78FAJw+fRqFhYVo2LChyrJzc3NhZWVVoXqI6P8xEBFRhRgbG6N+/fqVfr2W1pMz9eKpU0L5+fkqbQwNDZ+7jF69esHZ2RkrV66Eg4MDioqK0LRpU+Tl5VW6rufR1dWVflYoFAAghbCKyM7OBgCsXLkSPj4+KvO0tbWlnwMDAzFx4kQsW7YMsbGx8PT0hKenp7QMbW1tJCYmqrwGAExMTCpcExE9wTFERFRl3N3dcf36ddy6dUuadvjwYZU21tbWAKDS5tnL45s1a4b4+PhS13H37l1cuHAB06dPR7du3eDu7i4Nti6mp6cHANKYo9K4ublBT08PBw8elKbl5+fj2LFj8PDweM5WVp6trS0cHBxw5coV1K9fX+VRPBAcAHr37o3Hjx8jLi4OsbGxCAwMlOa1bNkShYWFSE9PL7EMOzu7aqmbSA54hIiIKiQ3Nxepqakq03R0dFCnTh34+fmhYcOGCAoKQnh4OLKysvDZZ5+ptK1fvz4cHR0xc+ZMfPHFF7h48SK++uorlTbTpk2Dp6cn3n//fYwdOxZ6enrYs2cPBgwYAEtLS1hZWSEqKgr29va4du0aPvnkE5XX29jYwNDQEHFxcXjttddgYGBQ4pJ7Y2NjjBs3DlOnToWlpSWcnJywYMECPHz4ECNHjnzpfkpJSSkR9Bo0aIBZs2Zh4sSJUCqV6NGjB3Jzc3H8+HHcv38fISEhUm19+vTBjBkzcP78eQwePFhaRsOGDREYGIhhw4bhq6++QsuWLXH79m3Ex8ejWbNmCAgIeOnaiWRJ3YOYiKj2CAoKEgBKPBo1aiS1uXDhgmjfvr3Q09MTDRs2FHFxcSUGGf/555/C09NTGBgYiA4dOogNGzaoDKoWQoi9e/eKtm3bCn19fWFubi78/f2lgdg7d+4U7u7uQl9fXzRr1kzs3bu3xDpWrlwpHB0dhZaWlujUqZNUf/GgaiGEePTokZgwYYKoU6eO0NfXF+3atRNHjx6V5pdnAHhpSusjAOLAgQNCCCFiYmJEixYthJ6enrCwsBAdO3YUv/76q8oytm3bJgCIjh07llh+Xl6eCA0NFS4uLkJXV1fY29uLvn37ilOnTgkhOKiaqDIUQjx1Ip+IiIhIhjiGiIiIiGSPgYiIiIhkj4GIiIiIZI+BiIiIiGSPgYiIiIhkj4GIiIiIZI+BiIiIiGSPgYiIiIhkj4GIiIiIZI+BiIiIiGSPgYiIiIhkj4GIiIiIZO//AFptRVJPPi0UAAAAAElFTkSuQmCC", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sns.countplot(x='education_level', hue='default_payment_next_month', data=data)\n", "plt.title('Education Level by Default Payment Next Month')\n", "plt.xlabel('Education Level')\n", "plt.ylabel('Count')\n", "plt.show()\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kelihatan bahwa kebanyakan client dari bank adalah lulusan \"graduate school\" dan \"university level\". Ini artinya proporsi dari populasi yang memiliki sarjana \"S1\" atau lebih tinggi sangat representatif dari semua client. Juga kelihatan bahwa yang jika pendidikan client lebih tinggi, maka mereka punya kemunkinan untuk bayar balik utang itu juga lebih tinggi. " ] }, { "cell_type": "code", "execution_count": 355, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sns.pairplot(data, vars=['limit_balance', 'age', 'marital_status', 'education_level', 'sex'], hue='default_payment_next_month')\n", "#plt.title('Pair Plot of Selected Features')\n", "plt.show()\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Disini kita mau lihat jika ada linear relationship antar variable 'age', 'marital_status', dan 'education_level'. Kelihatan data memiliki variable yang sangat luas dengan yang discrete dan continuous, jadi susah diprediksi apa saja yang memiliki linear relationship dari plot tersebut." ] }, { "cell_type": "code", "execution_count": 356, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sns.scatterplot(x='limit_balance', y='bill_amt_1', hue='default_payment_next_month', data=data)\n", "plt.title('Limit Balance vs Bill Amount 1')\n", "plt.xlabel('Limit Balance')\n", "plt.ylabel('Bill Amount 1')\n", "plt.show()\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Antar variable Limit balance data Bill_amount pertama, bisa kelihatan linear relationship yang positif antar client yang harus utang pembayaran bulan itu dengan total limit balance yang mereka punya di akun mereka. Bisa dibilang bahwa kebanyakan client cenderung memiliki limit baalance dibawah 500000, dan bill_amount mereka naik proporsional jika limit_balance juga lebih tinggi. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Split Train Test" ] }, { "cell_type": "code", "execution_count": 357, "metadata": {}, "outputs": [], "source": [ "from sklearn.model_selection import train_test_split\n", "\n", "# Define features and target variable\n", "X = data.drop('default_payment_next_month', axis=1)\n", "y = data['default_payment_next_month'] #yang di prediksi\n", "\n", "# Split data into training and testing sets\n", "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=50)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Goal kita untuk mencoba Split-train test itu untuk mengevaluasi skala \"performance\" dari model yang kita gunakan. Kita mau lihat beberapa baik model kita yang baru di-train akan reaksi terhadap data yang baru. Dengan test tersebut, kita akan bisa mengolah data dengan detecting over/underfitting atau menyatakan parameter di dalam training set. \n", "\n", "Kita juga mengunakan test size = 0.2, artinya 80% dari data akan digunakan untuk model training, dan 20% untuk cek \"performance\"." ] }, { "cell_type": "code", "execution_count": 358, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
limit_balancesexeducation_levelmarital_statusagepay_0pay_2pay_3pay_4pay_5...bill_amt_3bill_amt_4bill_amt_5bill_amt_6pay_amt_1pay_amt_2pay_amt_3pay_amt_4pay_amt_5pay_amt_6
21290000.011156.00.00.00.00.00.0...415700.0232732.0220460.0224780.010013.08501.015018.010001.07997.07624.0
1388260000.012228.00.00.02.00.00.0...163036.0159348.0160198.0104389.0107000.00.05000.05000.06000.060000.0
56760000.012253.00.00.00.00.00.0...57432.027126.027579.029085.03000.04000.02000.02000.02000.02000.0
1594150000.022223.0-2.0-1.0-1.00.00.0...151996.0152753.0153844.0151252.010096.0156292.04700.05019.05300.05002.0
1017130000.012228.02.02.00.00.00.0...70184.08518.011296.06514.00.03000.02000.03000.02000.00.0
..................................................................
201450000.021227.03.02.02.07.07.0...300.0300.0300.0300.00.00.00.00.00.00.0
215790000.023168.0-2.0-2.0-2.0-2.0-1.0...1000.01000.01052.069237.00.01000.01000.01052.071062.03000.0
1931150000.023148.02.02.02.02.02.0...62650.059255.045983.052986.05950.00.010000.00.020000.00.0
1504160000.013238.01.0-2.0-2.0-1.00.0...0.0700.0700.00.00.00.0700.00.00.00.0
171210000.021227.00.00.00.02.02.0...10255.09389.08345.08572.01400.02500.0500.00.0500.02000.0
\n", "

2372 rows × 23 columns

\n", "
" ], "text/plain": [ " limit_balance sex education_level marital_status age pay_0 pay_2 \\\n", "21 290000.0 1 1 1 56.0 0.0 0.0 \n", "1388 260000.0 1 2 2 28.0 0.0 0.0 \n", "567 60000.0 1 2 2 53.0 0.0 0.0 \n", "1594 150000.0 2 2 2 23.0 -2.0 -1.0 \n", "1017 130000.0 1 2 2 28.0 2.0 2.0 \n", "... ... ... ... ... ... ... ... \n", "2014 50000.0 2 1 2 27.0 3.0 2.0 \n", "2157 90000.0 2 3 1 68.0 -2.0 -2.0 \n", "1931 150000.0 2 3 1 48.0 2.0 2.0 \n", "1504 160000.0 1 3 2 38.0 1.0 -2.0 \n", "1712 10000.0 2 1 2 27.0 0.0 0.0 \n", "\n", " pay_3 pay_4 pay_5 ... bill_amt_3 bill_amt_4 bill_amt_5 \\\n", "21 0.0 0.0 0.0 ... 415700.0 232732.0 220460.0 \n", "1388 2.0 0.0 0.0 ... 163036.0 159348.0 160198.0 \n", "567 0.0 0.0 0.0 ... 57432.0 27126.0 27579.0 \n", "1594 -1.0 0.0 0.0 ... 151996.0 152753.0 153844.0 \n", "1017 0.0 0.0 0.0 ... 70184.0 8518.0 11296.0 \n", "... ... ... ... ... ... ... ... \n", "2014 2.0 7.0 7.0 ... 300.0 300.0 300.0 \n", "2157 -2.0 -2.0 -1.0 ... 1000.0 1000.0 1052.0 \n", "1931 2.0 2.0 2.0 ... 62650.0 59255.0 45983.0 \n", "1504 -2.0 -1.0 0.0 ... 0.0 700.0 700.0 \n", "1712 0.0 2.0 2.0 ... 10255.0 9389.0 8345.0 \n", "\n", " bill_amt_6 pay_amt_1 pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 \\\n", "21 224780.0 10013.0 8501.0 15018.0 10001.0 7997.0 \n", "1388 104389.0 107000.0 0.0 5000.0 5000.0 6000.0 \n", "567 29085.0 3000.0 4000.0 2000.0 2000.0 2000.0 \n", "1594 151252.0 10096.0 156292.0 4700.0 5019.0 5300.0 \n", "1017 6514.0 0.0 3000.0 2000.0 3000.0 2000.0 \n", "... ... ... ... ... ... ... \n", "2014 300.0 0.0 0.0 0.0 0.0 0.0 \n", "2157 69237.0 0.0 1000.0 1000.0 1052.0 71062.0 \n", "1931 52986.0 5950.0 0.0 10000.0 0.0 20000.0 \n", "1504 0.0 0.0 0.0 700.0 0.0 0.0 \n", "1712 8572.0 1400.0 2500.0 500.0 0.0 500.0 \n", "\n", " pay_amt_6 \n", "21 7624.0 \n", "1388 60000.0 \n", "567 2000.0 \n", "1594 5002.0 \n", "1017 0.0 \n", "... ... \n", "2014 0.0 \n", "2157 3000.0 \n", "1931 0.0 \n", "1504 0.0 \n", "1712 2000.0 \n", "\n", "[2372 rows x 23 columns]" ] }, "execution_count": 358, "metadata": {}, "output_type": "execute_result" } ], "source": [ "X_train" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Ini dari bagian X-train yang akan digunakan oleh machine learning. " ] }, { "cell_type": "code", "execution_count": 359, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "21 0\n", "1388 0\n", "567 0\n", "1594 0\n", "1017 1\n", " ..\n", "2014 1\n", "2157 0\n", "1931 1\n", "1504 0\n", "1712 0\n", "Name: default_payment_next_month, Length: 2372, dtype: int64" ] }, "execution_count": 359, "metadata": {}, "output_type": "execute_result" } ], "source": [ "y_train" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Ini dari kolom \"default_payment_next_month\" yang akan di-train dengan Machine Learning." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Handling Missing Values" ] }, { "cell_type": "code", "execution_count": 360, "metadata": {}, "outputs": [], "source": [ "X_train.dropna(inplace=True)" ] }, { "cell_type": "code", "execution_count": 361, "metadata": {}, "outputs": [], "source": [ "X_test.dropna(inplace=True)" ] }, { "cell_type": "code", "execution_count": 362, "metadata": {}, "outputs": [], "source": [ "y_train.dropna(inplace=True)" ] }, { "cell_type": "code", "execution_count": 363, "metadata": {}, "outputs": [], "source": [ "y_test.dropna(inplace=True)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Tadi kita cek tidak ada missing values di keseluruhan data, tapi kita cuman mau cek lagi jika ada missing values yang perlu di-drop lagi. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Handling Outliers" ] }, { "cell_type": "code", "execution_count": 364, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Categorical Columns: []\n", "Numerical Columns: ['limit_balance', 'sex', 'education_level', 'marital_status', 'age', 'pay_0', 'pay_2', 'pay_3', 'pay_4', 'pay_5', 'pay_6', 'bill_amt_1', 'bill_amt_2', 'bill_amt_3', 'bill_amt_4', 'bill_amt_5', 'bill_amt_6', 'pay_amt_1', 'pay_amt_2', 'pay_amt_3', 'pay_amt_4', 'pay_amt_5', 'pay_amt_6']\n" ] } ], "source": [ "# Selecting numerical and categorical columns\n", "num_columns = X_train.select_dtypes(include=['int64', 'float64']).columns.tolist()\n", "cat_columns = X_train.select_dtypes(include=['object', 'category']).columns.tolist()\n", "\n", "print('Categorical Columns: ', cat_columns)\n", "print('Numerical Columns: ', num_columns)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Untuk handle outliers, kita mau bagi dataset menjadi dua tipe kolom numerikal dan kategorikal. Ini untuk memastikan bisa diolah dengan training model dengan technique feature selection yang benar. Data yang berisi dengan integer atau float akan dimasukan ke kolom list num_columns, dan data dengan object atau category sebagai tipe akan dimasukan sebagai cat_columns. " ] }, { "cell_type": "code", "execution_count": 365, "metadata": {}, "outputs": [], "source": [ "# Making data and columns for normal distribution\n", "data_normal = []\n", "column_normal = []\n", "\n", "# Making data and columns for skewed distribution\n", "data_skewed = []\n", "column_skewed = []\n", "\n", "# For loop in every numerical column to filter the data distribution into either normally distributed or skewed columns\n", "for num in num_columns:\n", " skewness = X_train[num].skew()\n", "\n", " # If the data is normally distributed\n", " if skewness <= 0.5 and skewness >= -0.5:\n", " column_normal.append(num)\n", " data_normal.append([num, skewness])\n", "\n", " # If the data has low negative skewness\n", " elif skewness < -1:\n", " column_skewed.append(num)\n", " data_skewed.append([num, skewness, 'high'])\n", "\n", " # If the data has low positive skewness\n", " elif skewness > 1:\n", " column_skewed.append(num)\n", " data_skewed.append([num, skewness, 'high'])\n", "\n", " # If the data has moderate negative skewness\n", " elif skewness <= -0.5 and skewness > -1:\n", " column_skewed.append(num)\n", " data_skewed.append([num, skewness, 'low'])\n", "\n", " # If the data has moderate positive skewness\n", " elif skewness >= 0.5 and skewness < 1:\n", " column_skewed.append(num)\n", " data_skewed.append([num, skewness, 'low'])\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita mau perbedakan variable yang memiliki skew normal dan yang tidak normal." ] }, { "cell_type": "code", "execution_count": 366, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
normal_distributionskewness
0sex-0.451482
1marital_status-0.041328
\n", "
" ], "text/plain": [ " normal_distribution skewness\n", "0 sex -0.451482\n", "1 marital_status -0.041328" ] }, "execution_count": 366, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Showing normally distributed columns\n", "pd.DataFrame(data=data_normal, columns=['normal_distribution', 'skewness'])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kelihatan cuman sex dan marital_status memiliki skew yang normal." ] }, { "cell_type": "code", "execution_count": 367, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
skewed_distributionskewnessrate
0limit_balance1.013432high
1education_level0.913215low
2age0.757115low
3pay_00.910223low
4pay_20.852870low
5pay_30.970325low
6pay_41.153104high
7pay_51.079227high
8pay_61.011001high
9bill_amt_12.421285high
10bill_amt_22.438696high
11bill_amt_32.625471high
12bill_amt_42.508942high
13bill_amt_52.464289high
14bill_amt_62.503029high
15pay_amt_112.116671high
16pay_amt_228.106058high
17pay_amt_38.051309high
18pay_amt_48.888920high
19pay_amt_512.525019high
20pay_amt_610.277112high
\n", "
" ], "text/plain": [ " skewed_distribution skewness rate\n", "0 limit_balance 1.013432 high\n", "1 education_level 0.913215 low\n", "2 age 0.757115 low\n", "3 pay_0 0.910223 low\n", "4 pay_2 0.852870 low\n", "5 pay_3 0.970325 low\n", "6 pay_4 1.153104 high\n", "7 pay_5 1.079227 high\n", "8 pay_6 1.011001 high\n", "9 bill_amt_1 2.421285 high\n", "10 bill_amt_2 2.438696 high\n", "11 bill_amt_3 2.625471 high\n", "12 bill_amt_4 2.508942 high\n", "13 bill_amt_5 2.464289 high\n", "14 bill_amt_6 2.503029 high\n", "15 pay_amt_1 12.116671 high\n", "16 pay_amt_2 28.106058 high\n", "17 pay_amt_3 8.051309 high\n", "18 pay_amt_4 8.888920 high\n", "19 pay_amt_5 12.525019 high\n", "20 pay_amt_6 10.277112 high" ] }, "execution_count": 367, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Showing skewed columns\n", "pd.DataFrame(data=data_skewed, columns=['skewed_distribution', 'skewness', 'rate'])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Banyak variable di dataset yang memiliki skew yang tinggi daripada yang rendah. Ini artinya dataset memiliki skew yang cenderung positif. " ] }, { "cell_type": "code", "execution_count": 368, "metadata": {}, "outputs": [], "source": [ "# Capping Method for Normal Distribution\n", "winsorizer_normal = Winsorizer(capping_method='gaussian',\n", " tail='both',\n", " fold=3,\n", " variables=column_normal,\n", " missing_values='ignore')\n", "\n", "# Fit & Transforming X_train\n", "X_train_capped = winsorizer_normal.fit_transform(X_train)\n", "\n", "# Transforming X_test\n", "X_test_capped = winsorizer_normal.transform(X_test)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita mau coba \"cap\" outlier yang memiliki skew normal. Ini jadi dataset tidak akan terlalu sensitif terhadap outlier yang akan diubah. " ] }, { "cell_type": "code", "execution_count": 369, "metadata": {}, "outputs": [], "source": [ "# Capping Method for Skewed Distribution\n", "winsorizer_skewed = Winsorizer(capping_method='iqr',\n", " tail='both',\n", " fold=3,\n", " variables=column_skewed)\n", "\n", "# Fit & Transforming X_train\n", "X_train_capped = winsorizer_skewed.fit_transform(X_train_capped)\n", "\n", "# Transforming X_test\n", "X_test_capped = winsorizer_skewed.transform(X_test_capped)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita mau coba \"cap\" outlier yang memiliki skew tidak normal. Ini jadi dataset tidak akan terlalu sensitif terhadap outlier yang akan diubah. " ] }, { "cell_type": "code", "execution_count": 370, "metadata": {}, "outputs": [], "source": [ "# Plot Distribution Comparison\n", "def outlier_handling_plot_comparison(df_before, df_after, variable):\n", "\n", " # Figure Size, and Super Title based on variable\n", " fig, axes = plt.subplots(2, 2, figsize=(15, 10))\n", " fig.suptitle(f'{variable} - Distribution Before and After Outlier Handling')\n", "\n", " # Plot Histogram Before\n", " sns.histplot(df_before[variable], bins=30, ax=axes[0, 0], color='orange')\n", " axes[0, 0].set_title('Histogram Before')\n", "\n", " # Plot Boxplot Before\n", " sns.boxplot(y=df_before[variable], ax=axes[1, 0])\n", " axes[1, 0].set_title('Boxplot Before')\n", "\n", " # Plot Histogram After\n", " sns.histplot(df_after[variable], bins=30, ax=axes[0, 1], color='orange')\n", " axes[0, 1].set_title('Histogram After')\n", "\n", " # Plot Boxplot After\n", " sns.boxplot(y=df_after[variable], ax=axes[1, 1])\n", " axes[1, 1].set_title('Boxplot After')\n", "\n", " plt.tight_layout(rect=[0, 0.03, 1, 0.95])\n", " plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Ini kita mau coba illustrasi variable apa saja yang telah diubah dengan beberapa \"capping\" metode. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Feature Engineering" ] }, { "cell_type": "code", "execution_count": 371, "metadata": {}, "outputs": [], "source": [ "\n", "list_num_col = ['limit_balance', 'age', 'pay_0', 'pay_2', 'pay_3', 'pay_4', 'pay_5', 'pay_6', \n", " 'bill_amt_1', 'bill_amt_2', 'bill_amt_3', 'bill_amt_4', 'bill_amt_5', 'bill_amt_6', \n", " 'pay_amt_1', 'pay_amt_2', 'pay_amt_3', 'pay_amt_4', 'pay_amt_5', 'pay_amt_6']\n", "list_cat_col = ['sex', 'education_level', 'marital_status']\n", "\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Untuk di feature engineering, kita mau define beberapa variable category secara manual karena beberapa variable kategori sudah di encode. Untuk mempermudah machine learning nanti, kita akan coba perbedaan variable numerikal dan kategorikal. " ] }, { "cell_type": "code", "execution_count": 372, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
limit_balancesexeducation_levelmarital_statusagepay_0pay_2pay_3pay_4pay_5...bill_amt_3bill_amt_4bill_amt_5bill_amt_6pay_amt_1pay_amt_2pay_amt_3pay_amt_4pay_amt_5pay_amt_6
21290000.0111.056.00.00.00.00.00.0...240879.75223294.75207163.0204697.2510013.008501.0015018.010001.07997.007624.0
1388260000.0122.028.00.00.02.00.00.0...163036.00159348.00160198.0104389.0017120.750.005000.05000.06000.0015427.0
56760000.0122.053.00.00.00.00.00.0...57432.0027126.0027579.029085.003000.004000.002000.02000.02000.002000.0
1594150000.0222.023.0-2.0-1.0-1.00.00.0...151996.00152753.00153844.0151252.0010096.0017215.254700.05019.05300.005002.0
1017130000.0122.028.02.02.00.00.00.0...70184.008518.0011296.06514.000.003000.002000.03000.02000.000.0
..................................................................
201450000.0212.027.03.02.02.03.03.0...300.00300.00300.0300.000.000.000.00.00.000.0
215790000.0231.068.0-2.0-2.0-2.0-2.0-1.0...1000.001000.001052.069237.000.001000.001000.01052.014871.253000.0
1931150000.0231.048.02.02.02.02.02.0...62650.0059255.0045983.052986.005950.000.0010000.00.014871.250.0
1504160000.0132.038.01.0-2.0-2.0-1.00.0...0.00700.00700.00.000.000.00700.00.00.000.0
171210000.0212.027.00.00.00.02.02.0...10255.009389.008345.08572.001400.002500.00500.00.0500.002000.0
\n", "

2372 rows × 23 columns

\n", "
" ], "text/plain": [ " limit_balance sex education_level marital_status age pay_0 pay_2 \\\n", "21 290000.0 1 1 1.0 56.0 0.0 0.0 \n", "1388 260000.0 1 2 2.0 28.0 0.0 0.0 \n", "567 60000.0 1 2 2.0 53.0 0.0 0.0 \n", "1594 150000.0 2 2 2.0 23.0 -2.0 -1.0 \n", "1017 130000.0 1 2 2.0 28.0 2.0 2.0 \n", "... ... ... ... ... ... ... ... \n", "2014 50000.0 2 1 2.0 27.0 3.0 2.0 \n", "2157 90000.0 2 3 1.0 68.0 -2.0 -2.0 \n", "1931 150000.0 2 3 1.0 48.0 2.0 2.0 \n", "1504 160000.0 1 3 2.0 38.0 1.0 -2.0 \n", "1712 10000.0 2 1 2.0 27.0 0.0 0.0 \n", "\n", " pay_3 pay_4 pay_5 ... bill_amt_3 bill_amt_4 bill_amt_5 \\\n", "21 0.0 0.0 0.0 ... 240879.75 223294.75 207163.0 \n", "1388 2.0 0.0 0.0 ... 163036.00 159348.00 160198.0 \n", "567 0.0 0.0 0.0 ... 57432.00 27126.00 27579.0 \n", "1594 -1.0 0.0 0.0 ... 151996.00 152753.00 153844.0 \n", "1017 0.0 0.0 0.0 ... 70184.00 8518.00 11296.0 \n", "... ... ... ... ... ... ... ... \n", "2014 2.0 3.0 3.0 ... 300.00 300.00 300.0 \n", "2157 -2.0 -2.0 -1.0 ... 1000.00 1000.00 1052.0 \n", "1931 2.0 2.0 2.0 ... 62650.00 59255.00 45983.0 \n", "1504 -2.0 -1.0 0.0 ... 0.00 700.00 700.0 \n", "1712 0.0 2.0 2.0 ... 10255.00 9389.00 8345.0 \n", "\n", " bill_amt_6 pay_amt_1 pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 \\\n", "21 204697.25 10013.00 8501.00 15018.0 10001.0 7997.00 \n", "1388 104389.00 17120.75 0.00 5000.0 5000.0 6000.00 \n", "567 29085.00 3000.00 4000.00 2000.0 2000.0 2000.00 \n", "1594 151252.00 10096.00 17215.25 4700.0 5019.0 5300.00 \n", "1017 6514.00 0.00 3000.00 2000.0 3000.0 2000.00 \n", "... ... ... ... ... ... ... \n", "2014 300.00 0.00 0.00 0.0 0.0 0.00 \n", "2157 69237.00 0.00 1000.00 1000.0 1052.0 14871.25 \n", "1931 52986.00 5950.00 0.00 10000.0 0.0 14871.25 \n", "1504 0.00 0.00 0.00 700.0 0.0 0.00 \n", "1712 8572.00 1400.00 2500.00 500.0 0.0 500.00 \n", "\n", " pay_amt_6 \n", "21 7624.0 \n", "1388 15427.0 \n", "567 2000.0 \n", "1594 5002.0 \n", "1017 0.0 \n", "... ... \n", "2014 0.0 \n", "2157 3000.0 \n", "1931 0.0 \n", "1504 0.0 \n", "1712 2000.0 \n", "\n", "[2372 rows x 23 columns]" ] }, "execution_count": 372, "metadata": {}, "output_type": "execute_result" } ], "source": [ "X_train_capped" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Coba lihat data dari X_train yang sudah diubah outlier-nya. " ] }, { "cell_type": "code", "execution_count": 373, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
limit_balanceagepay_0pay_2pay_3pay_4pay_5pay_6bill_amt_1bill_amt_2bill_amt_3bill_amt_4bill_amt_5bill_amt_6pay_amt_1pay_amt_2pay_amt_3pay_amt_4pay_amt_5pay_amt_6
21290000.056.00.00.00.00.00.00.0222000.0226917.0240879.75223294.75207163.0204697.2510013.008501.0015018.010001.07997.07624.0
1388260000.028.00.00.02.00.00.00.0149814.0184419.0163036.00159348.00160198.0104389.0017120.750.005000.05000.06000.015427.0
56760000.053.00.00.00.00.00.00.056765.057849.057432.0027126.0027579.029085.003000.004000.002000.02000.02000.02000.0
1594150000.023.0-2.0-1.0-1.00.00.00.027414.010053.0151996.00152753.00153844.0151252.0010096.0017215.254700.05019.05300.05002.0
1017130000.028.02.02.00.00.00.00.070952.069264.070184.008518.0011296.06514.000.003000.002000.03000.02000.00.0
\n", "
" ], "text/plain": [ " limit_balance age pay_0 pay_2 pay_3 pay_4 pay_5 pay_6 \\\n", "21 290000.0 56.0 0.0 0.0 0.0 0.0 0.0 0.0 \n", "1388 260000.0 28.0 0.0 0.0 2.0 0.0 0.0 0.0 \n", "567 60000.0 53.0 0.0 0.0 0.0 0.0 0.0 0.0 \n", "1594 150000.0 23.0 -2.0 -1.0 -1.0 0.0 0.0 0.0 \n", "1017 130000.0 28.0 2.0 2.0 0.0 0.0 0.0 0.0 \n", "\n", " bill_amt_1 bill_amt_2 bill_amt_3 bill_amt_4 bill_amt_5 bill_amt_6 \\\n", "21 222000.0 226917.0 240879.75 223294.75 207163.0 204697.25 \n", "1388 149814.0 184419.0 163036.00 159348.00 160198.0 104389.00 \n", "567 56765.0 57849.0 57432.00 27126.00 27579.0 29085.00 \n", "1594 27414.0 10053.0 151996.00 152753.00 153844.0 151252.00 \n", "1017 70952.0 69264.0 70184.00 8518.00 11296.0 6514.00 \n", "\n", " pay_amt_1 pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 pay_amt_6 \n", "21 10013.00 8501.00 15018.0 10001.0 7997.0 7624.0 \n", "1388 17120.75 0.00 5000.0 5000.0 6000.0 15427.0 \n", "567 3000.00 4000.00 2000.0 2000.0 2000.0 2000.0 \n", "1594 10096.00 17215.25 4700.0 5019.0 5300.0 5002.0 \n", "1017 0.00 3000.00 2000.0 3000.0 2000.0 0.0 " ] }, "execution_count": 373, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Splitting the train and test features into categorical and numerical columns\n", "X_train_num = X_train_capped[list_num_col]\n", "X_train_cat = X_train_capped[list_cat_col]\n", "\n", "X_test_num = X_test_capped[list_num_col]\n", "X_test_cat = X_test_capped[list_cat_col]\n", "\n", "X_train_num.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita mau assign numerikal dan kategorikal kolom kepada variabble baru untuk diuji korelasi. " ] }, { "cell_type": "code", "execution_count": 374, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Column NameCorrelation CoefficientP-valueCorrelation
0sex-0.0263330.199759Not Significant
1education_level0.0632560.001185Significant
2marital_status-0.0328350.107643Not Significant
\n", "
" ], "text/plain": [ " Column Name Correlation Coefficient P-value Correlation\n", "0 sex -0.026333 0.199759 Not Significant\n", "1 education_level 0.063256 0.001185 Significant\n", "2 marital_status -0.032835 0.107643 Not Significant" ] }, "execution_count": 374, "metadata": {}, "output_type": "execute_result" } ], "source": [ "#Finding the correlation between categorical columns and Y Train using Kendall Tau's correlation\n", "p_values = []\n", "interpretation = []\n", "cols = []\n", "corr = []\n", "selected_cat_cols = []\n", "\n", "for col in X_train_cat.columns:\n", " corr_coef, p_value = kendalltau(X_train_cat[col], y_train)\n", "\n", " p_values.append(p_value)\n", " cols.append(col)\n", " corr.append(corr_coef)\n", "\n", " if p_value < 0.05:\n", " interpretation.append('Significant')\n", " selected_cat_cols.append(col)\n", " else :\n", " interpretation.append('Not Significant')\n", "\n", "pd.DataFrame({'Column Name':cols,\n", " 'Correlation Coefficient' : corr,\n", " 'P-value':p_values,\n", " 'Correlation': interpretation })" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Gunakan KendallTau test terhadap kategorikal kolom untuk melihat p-value yang signifikan." ] }, { "cell_type": "code", "execution_count": 375, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Column NameCorrelation CoefficientP-valueCorrelation
0limit_balance-0.1704286.420760e-17Significant
1age0.0291201.562476e-01Not Significant
2pay_00.3587165.899447e-73Significant
3pay_20.2376268.379995e-32Significant
4pay_30.2229754.151099e-28Significant
5pay_40.2531775.201433e-36Significant
6pay_50.2518011.260071e-35Significant
7pay_60.2306255.259622e-30Significant
8bill_amt_1-0.0082146.892859e-01Not Significant
9bill_amt_20.0037058.568704e-01Not Significant
10bill_amt_30.0053527.944745e-01Not Significant
11bill_amt_40.0058197.769890e-01Not Significant
12bill_amt_50.0113195.816419e-01Not Significant
13bill_amt_60.0203263.223980e-01Not Significant
14pay_amt_1-0.1442921.655782e-12Significant
15pay_amt_2-0.1478314.627461e-13Significant
16pay_amt_3-0.1179318.375614e-09Significant
17pay_amt_4-0.1289782.888231e-10Significant
18pay_amt_5-0.0955803.112395e-06Significant
19pay_amt_6-0.1370282.058593e-11Significant
\n", "
" ], "text/plain": [ " Column Name Correlation Coefficient P-value Correlation\n", "0 limit_balance -0.170428 6.420760e-17 Significant\n", "1 age 0.029120 1.562476e-01 Not Significant\n", "2 pay_0 0.358716 5.899447e-73 Significant\n", "3 pay_2 0.237626 8.379995e-32 Significant\n", "4 pay_3 0.222975 4.151099e-28 Significant\n", "5 pay_4 0.253177 5.201433e-36 Significant\n", "6 pay_5 0.251801 1.260071e-35 Significant\n", "7 pay_6 0.230625 5.259622e-30 Significant\n", "8 bill_amt_1 -0.008214 6.892859e-01 Not Significant\n", "9 bill_amt_2 0.003705 8.568704e-01 Not Significant\n", "10 bill_amt_3 0.005352 7.944745e-01 Not Significant\n", "11 bill_amt_4 0.005819 7.769890e-01 Not Significant\n", "12 bill_amt_5 0.011319 5.816419e-01 Not Significant\n", "13 bill_amt_6 0.020326 3.223980e-01 Not Significant\n", "14 pay_amt_1 -0.144292 1.655782e-12 Significant\n", "15 pay_amt_2 -0.147831 4.627461e-13 Significant\n", "16 pay_amt_3 -0.117931 8.375614e-09 Significant\n", "17 pay_amt_4 -0.128978 2.888231e-10 Significant\n", "18 pay_amt_5 -0.095580 3.112395e-06 Significant\n", "19 pay_amt_6 -0.137028 2.058593e-11 Significant" ] }, "execution_count": 375, "metadata": {}, "output_type": "execute_result" } ], "source": [ "#Finding the correlation between numerical columns and Y Train using pearsonr and spearmanr correlation\n", "p_values = []\n", "interpretation = []\n", "cols = []\n", "corr = []\n", "selected_num_cols = []\n", "\n", "for col in X_train_num.columns:\n", " if abs(X_train_num[col].skew()) < 0.5:\n", " #For Normally Distributed Columns\n", " corr_coef, p_value = pearsonr(X_train_num[col], y_train)\n", "\n", " p_values.append(p_value)\n", " cols.append(col)\n", " corr.append(corr_coef)\n", "\n", " if p_value < 0.05:\n", " interpretation.append('Significant')\n", " selected_num_cols.append(col)\n", " else :\n", " interpretation.append('Not Significant')\n", " else:\n", " #For Skewed Columns\n", " corr_coef, p_value = spearmanr(X_train_num[col], y_train)\n", "\n", " p_values.append(p_value)\n", " cols.append(col)\n", " corr.append(corr_coef)\n", "\n", " if p_value < 0.05:\n", " interpretation.append('Significant')\n", " selected_num_cols.append(col)\n", " else :\n", " interpretation.append('Not Significant')\n", "\n", "pd.DataFrame({'Column Name':cols,\n", " 'Correlation Coefficient' : corr,\n", " 'P-value':p_values,\n", " 'Correlation': interpretation })" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Gunakan Pearson dan Spearman test terhadap numerikal kolom untuk melihat p-value yang signifikan." ] }, { "cell_type": "code", "execution_count": 376, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "['education_level']\n", "['limit_balance', 'pay_0', 'pay_2', 'pay_3', 'pay_4', 'pay_5', 'pay_6', 'pay_amt_1', 'pay_amt_2', 'pay_amt_3', 'pay_amt_4', 'pay_amt_5', 'pay_amt_6']\n" ] } ], "source": [ "# Show selected columns based on the correlation test\n", "print(selected_cat_cols)\n", "print(selected_num_cols)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan ambil data yang signifikan dari kategorikal dan numerikal kolom untuk di scaling nanti. Ini akan jadi \"base\" data yang akan digunakan Machine Learning untuk memastikan \"fitting\" terhadap model yang akan dibuat. " ] }, { "cell_type": "code", "execution_count": 377, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
limit_balancepay_0pay_2pay_3pay_4pay_5pay_6pay_amt_1pay_amt_2pay_amt_3pay_amt_4pay_amt_5pay_amt_6
21290000.00.00.00.00.00.00.010013.008501.0015018.010001.07997.07624.0
1388260000.00.00.02.00.00.00.017120.750.005000.05000.06000.015427.0
56760000.00.00.00.00.00.00.03000.004000.002000.02000.02000.02000.0
1594150000.0-2.0-1.0-1.00.00.00.010096.0017215.254700.05019.05300.05002.0
1017130000.02.02.00.00.00.00.00.003000.002000.03000.02000.00.0
\n", "
" ], "text/plain": [ " limit_balance pay_0 pay_2 pay_3 pay_4 pay_5 pay_6 pay_amt_1 \\\n", "21 290000.0 0.0 0.0 0.0 0.0 0.0 0.0 10013.00 \n", "1388 260000.0 0.0 0.0 2.0 0.0 0.0 0.0 17120.75 \n", "567 60000.0 0.0 0.0 0.0 0.0 0.0 0.0 3000.00 \n", "1594 150000.0 -2.0 -1.0 -1.0 0.0 0.0 0.0 10096.00 \n", "1017 130000.0 2.0 2.0 0.0 0.0 0.0 0.0 0.00 \n", "\n", " pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 pay_amt_6 \n", "21 8501.00 15018.0 10001.0 7997.0 7624.0 \n", "1388 0.00 5000.0 5000.0 6000.0 15427.0 \n", "567 4000.00 2000.0 2000.0 2000.0 2000.0 \n", "1594 17215.25 4700.0 5019.0 5300.0 5002.0 \n", "1017 3000.00 2000.0 3000.0 2000.0 0.0 " ] }, "execution_count": 377, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Updating Numerical and Categorical Columns\n", "X_train_cat = X_train_cat[selected_cat_cols]\n", "X_train_num = X_train_num[selected_num_cols]\n", "\n", "X_test_cat = X_test_cat[selected_cat_cols]\n", "X_test_num = X_test_num[selected_num_cols]\n", "\n", "#Show first five data from the updated X_train\n", "X_train_num.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Letakan kategorikal dan numerikal kolom yang terpilih tadi ke variable baru untuk siap di-training. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Scaling" ] }, { "cell_type": "code", "execution_count": 378, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([[0.36842105, 0.4 , 0.4 , ..., 0.66426448, 0.53774901,\n", " 0.49419848],\n", " [0.32894737, 0.4 , 0.4 , ..., 0.33209903, 0.40346306,\n", " 1. ],\n", " [0.06578947, 0.4 , 0.4 , ..., 0.13283961, 0.13448769,\n", " 0.12964283],\n", " ...,\n", " [0.18421053, 0.8 , 0.8 , ..., 0. , 1. ,\n", " 0. ],\n", " [0.19736842, 0.6 , 0. , ..., 0. , 0. ,\n", " 0. ],\n", " [0. , 0.4 , 0.4 , ..., 0. , 0.03362192,\n", " 0.12964283]])" ] }, "execution_count": 378, "metadata": {}, "output_type": "execute_result" } ], "source": [ "#Initialize the MinMaxScaler\n", "scaler = MinMaxScaler()\n", "\n", "#Fit_transform for X_train, transform for X_test\n", "X_train_scaled = scaler.fit_transform(X_train_num)\n", "X_test_scaled = scaler.transform(X_test_num)\n", "\n", "X_train_scaled" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Untuk numerikal data, kita akan mengunakan MinMaxScaler daripada Standard atau Robust scaler karena kita mau coba simpan distribusi original data tanpa mengubah perbedaan \"range\" antar values yang telah ditampilkan. Juga tadi outlier sudah di \"cap\" dengan Winsoriser, jadi sekarang data numerikal yang akan di-scale akan lebih sensitif terhadap perbedaan values. Juga distribusi dari numerikal values juga tidak normal, jadi StandardScaler tidak sesuai untuk variasi yang terlalu skewed. " ] }, { "cell_type": "code", "execution_count": 379, "metadata": {}, "outputs": [], "source": [ "X_train_final = np.concatenate([X_train_cat, X_train_scaled], axis=1)\n", "X_test_final = np.concatenate([X_test_cat, X_test_scaled], axis=1)\n", "\n", "# Get the column names\n", "column_names = list(X_train_cat.columns) + list(X_train_num.columns)\n", "\n", "# Create DataFrames with column names\n", "X_train_final = pd.DataFrame(X_train_final, columns=column_names)\n", "X_test_final = pd.DataFrame(X_test_final, columns=column_names)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan taruh balik dari X_train dan X_test variables dengan yang sudah di \"scaled\" untuk membuat list baru di variable \"final\". X_train_final dan X_test_final akan digunakan untuk model yang akan diuji. " ] }, { "cell_type": "code", "execution_count": 380, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
education_levellimit_balancepay_0pay_2pay_3pay_4pay_5pay_6pay_amt_1pay_amt_2pay_amt_3pay_amt_4pay_amt_5pay_amt_6
01.00.3684210.40.40.40.40.40.40.5848460.4938060.9164300.6642640.5377490.494198
12.00.3289470.40.40.80.40.40.41.0000000.0000000.3051110.3320990.4034631.000000
22.00.0657890.40.40.40.40.40.40.1752260.2323520.1220440.1328400.1344880.129643
32.00.1842110.00.20.20.40.40.40.5896941.0000000.2868040.3333610.3563920.324237
42.00.1578950.80.80.40.40.40.40.0000000.1742640.1220440.1992590.1344880.000000
.............................................
23671.00.0526321.00.80.81.01.01.00.0000000.0000000.0000000.0000000.0000000.000000
23683.00.1052630.00.00.00.00.20.20.0000000.0580880.0610220.0698741.0000000.194464
23693.00.1842110.80.80.80.80.80.80.3475320.0000000.6102210.0000001.0000000.000000
23703.00.1973680.60.00.00.20.40.40.0000000.0000000.0427150.0000000.0000000.000000
23711.00.0000000.40.40.40.80.80.40.0817720.1452200.0305110.0000000.0336220.129643
\n", "

2372 rows × 14 columns

\n", "
" ], "text/plain": [ " education_level limit_balance pay_0 pay_2 pay_3 pay_4 pay_5 \\\n", "0 1.0 0.368421 0.4 0.4 0.4 0.4 0.4 \n", "1 2.0 0.328947 0.4 0.4 0.8 0.4 0.4 \n", "2 2.0 0.065789 0.4 0.4 0.4 0.4 0.4 \n", "3 2.0 0.184211 0.0 0.2 0.2 0.4 0.4 \n", "4 2.0 0.157895 0.8 0.8 0.4 0.4 0.4 \n", "... ... ... ... ... ... ... ... \n", "2367 1.0 0.052632 1.0 0.8 0.8 1.0 1.0 \n", "2368 3.0 0.105263 0.0 0.0 0.0 0.0 0.2 \n", "2369 3.0 0.184211 0.8 0.8 0.8 0.8 0.8 \n", "2370 3.0 0.197368 0.6 0.0 0.0 0.2 0.4 \n", "2371 1.0 0.000000 0.4 0.4 0.4 0.8 0.8 \n", "\n", " pay_6 pay_amt_1 pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 pay_amt_6 \n", "0 0.4 0.584846 0.493806 0.916430 0.664264 0.537749 0.494198 \n", "1 0.4 1.000000 0.000000 0.305111 0.332099 0.403463 1.000000 \n", "2 0.4 0.175226 0.232352 0.122044 0.132840 0.134488 0.129643 \n", "3 0.4 0.589694 1.000000 0.286804 0.333361 0.356392 0.324237 \n", "4 0.4 0.000000 0.174264 0.122044 0.199259 0.134488 0.000000 \n", "... ... ... ... ... ... ... ... \n", "2367 1.0 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 \n", "2368 0.2 0.000000 0.058088 0.061022 0.069874 1.000000 0.194464 \n", "2369 0.8 0.347532 0.000000 0.610221 0.000000 1.000000 0.000000 \n", "2370 0.4 0.000000 0.000000 0.042715 0.000000 0.000000 0.000000 \n", "2371 0.4 0.081772 0.145220 0.030511 0.000000 0.033622 0.129643 \n", "\n", "[2372 rows x 14 columns]" ] }, "execution_count": 380, "metadata": {}, "output_type": "execute_result" } ], "source": [ "X_train_final" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Cek lagi jika sudah di-concate dengan benar. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Model Definition" ] }, { "cell_type": "code", "execution_count": 381, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
LogisticRegression()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "LogisticRegression()" ] }, "execution_count": 381, "metadata": {}, "output_type": "execute_result" } ], "source": [ "lg = LogisticRegression()\n", "lg.fit(X_train_final, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan mengunkan Logistic Regression daripada Linear karena tipe data yang diujui, \"default\" atau tidak, itu adalah binary dengan dua outcome. Ini artinya output dari probability value itu antara 0 dan 1. " ] }, { "cell_type": "code", "execution_count": 382, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
KNeighborsClassifier(n_neighbors=11)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "KNeighborsClassifier(n_neighbors=11)" ] }, "execution_count": 382, "metadata": {}, "output_type": "execute_result" } ], "source": [ "knn = KNeighborsClassifier(n_neighbors=11)\n", "knn.fit(X_train_scaled, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan menggunakan KNN test untuk mengklasifikasi dataset yang ada jarak dari K, dan perbedakan klasifikasi dari satu data-point dengan yang lain di sekitarnya. Kita mau lihat nilai rata-rata weighted average dari jarak K value yang kita akan gunakan. " ] }, { "cell_type": "code", "execution_count": 383, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
SVC()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "SVC()" ] }, "execution_count": 383, "metadata": {}, "output_type": "execute_result" } ], "source": [ "svc = SVC()\n", "svc.fit(X_train_scaled, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "SVC akan digunakan untuk klasifisikan beberapa kelas dengan hyperplane dan support vectors. Kita juga akan coba uji dengan Regularization parameter (C) dan Kernel trick untuk lihat mana nilai akurasi, precision, dan recall yang paling sesuai untuk test tersebut. " ] }, { "cell_type": "code", "execution_count": 384, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
SVC()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "SVC()" ] }, "execution_count": 384, "metadata": {}, "output_type": "execute_result" } ], "source": [ "svm_non_scaled = SVC(kernel='rbf')\n", "svm_scaled = SVC(kernel='rbf')\n", "\n", "svm_non_scaled.fit(X_train, y_train)\n", "svm_scaled.fit(X_train_scaled, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Untuk test SVM_scaled, akan mengunakan RBF karena kita tidak tau data distribusi tersebut, juga kita asumsi data boundary tidak linear. " ] }, { "cell_type": "code", "execution_count": 385, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
SVC()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "SVC()" ] }, "execution_count": 385, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Model Training using different kernels\n", "\n", "svm_linear = SVC(kernel='linear')\n", "svm_poly = SVC(kernel='poly')\n", "svm_rbf = SVC(kernel='rbf')\n", "\n", "svm_linear.fit(X_train_scaled, y_train)\n", "svm_poly.fit(X_train_scaled, y_train)\n", "svm_rbf.fit(X_train_scaled, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan coba perbandingkan \"linear\", \"polynomial\", dan \"RBF\" kernel untuk lihat nilai test. " ] }, { "cell_type": "code", "execution_count": 386, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
SVC(C=500)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "SVC(C=500)" ] }, "execution_count": 386, "metadata": {}, "output_type": "execute_result" } ], "source": [ "svm_rbf_1 = SVC(kernel='rbf', C=0.1)\n", "svm_rbf_500 = SVC(kernel='rbf', C=500)\n", "\n", "svm_rbf_1.fit(X_train_scaled, y_train)\n", "svm_rbf_500.fit(X_train_scaled, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Gunakan C regularization untuk minimalisir classification error yang akan menjadi test \"overfitting\". " ] }, { "cell_type": "code", "execution_count": 387, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
SVC(C=500, gamma=100)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "SVC(C=500, gamma=100)" ] }, "execution_count": 387, "metadata": {}, "output_type": "execute_result" } ], "source": [ "svm_rbf_500_1 = SVC(kernel='rbf', C=500, gamma=0.1)\n", "svm_rbf_500_100 = SVC(kernel='rbf', C=500, gamma=100)\n", "\n", "svm_rbf_500_1.fit(X_train_scaled, y_train)\n", "svm_rbf_500_100.fit(X_train_scaled, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Coba Gamma test untuk melihat beberapa influence satu point akan terhadap keseluruhan test. " ] }, { "cell_type": "code", "execution_count": 388, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
DecisionTreeClassifier(max_depth=6, random_state=10)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "DecisionTreeClassifier(max_depth=6, random_state=10)" ] }, "execution_count": 388, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Training using Decision Tree\n", "from sklearn.tree import DecisionTreeClassifier\n", "\n", "model_dt = DecisionTreeClassifier(max_depth=6, random_state=10)\n", "model_dt.fit(X_train_scaled, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Gunakan Decision tree untuk mengillustrasi perbedaan dari data points yang telah diklasifikasikan oleh Machine Learning. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Model Training" ] }, { "cell_type": "code", "execution_count": 389, "metadata": {}, "outputs": [], "source": [ "y_pred_train_lg = lg.predict(X_train_final)\n", "y_pred_test_lg = lg.predict(X_test_final)" ] }, { "cell_type": "code", "execution_count": 390, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " precision recall f1-score support\n", "\n", " 0 0.83 0.97 0.90 1863\n", " 1 0.72 0.28 0.40 509\n", "\n", " accuracy 0.82 2372\n", " macro avg 0.77 0.62 0.65 2372\n", "weighted avg 0.81 0.82 0.79 2372\n", "\n", " precision recall f1-score support\n", "\n", " 0 0.84 0.97 0.90 467\n", " 1 0.75 0.31 0.44 126\n", "\n", " accuracy 0.83 593\n", " macro avg 0.79 0.64 0.67 593\n", "weighted avg 0.82 0.83 0.80 593\n", "\n" ] } ], "source": [ "# Model Evaluation - Train Set & Test Set\n", "\n", "print(classification_report(y_train, y_pred_train_lg))\n", "print(classification_report(y_test, y_pred_test_lg))" ] }, { "cell_type": "code", "execution_count": 391, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Train Set Evaluation:\n", "accuracy: 0.8217\n", "precision (weighted avg): 0.8067\n", "recall (weighted avg): 0.8217\n", "f1-score (weighted avg): 0.7893\n", "Test Set Evaluation:\n", "accuracy: 0.8314\n", "precision (weighted avg): 0.8202\n", "recall (weighted avg): 0.8314\n", "f1-score (weighted avg): 0.8025\n" ] } ], "source": [ "# Model Evaluation - Train Set\n", "train_report_dict = classification_report(y_train, y_pred_train_lg, output_dict=True)\n", "train_accuracy = accuracy_score(y_train, y_pred_train_lg)\n", "\n", "print(\"Train Set Evaluation:\")\n", "#print(classification_report(y_train, y_pred_train_lg))\n", "print(f\"accuracy: {train_accuracy:.4f}\")\n", "print(f\"precision (weighted avg): {train_report_dict['weighted avg']['precision']:.4f}\")\n", "print(f\"recall (weighted avg): {train_report_dict['weighted avg']['recall']:.4f}\")\n", "print(f\"f1-score (weighted avg): {train_report_dict['weighted avg']['f1-score']:.4f}\")\n", "\n", "# Model Evaluation - Test Set\n", "test_report_dict = classification_report(y_test, y_pred_test_lg, output_dict=True)\n", "test_accuracy = accuracy_score(y_test, y_pred_test_lg)\n", "\n", "print(\"Test Set Evaluation:\")\n", "#print(classification_report(y_test, y_pred_test_lg))\n", "print(f\"accuracy: {test_accuracy:.4f}\")\n", "print(f\"precision (weighted avg): {test_report_dict['weighted avg']['precision']:.4f}\")\n", "print(f\"recall (weighted avg): {test_report_dict['weighted avg']['recall']:.4f}\")\n", "print(f\"f1-score (weighted avg): {test_report_dict['weighted avg']['f1-score']:.4f}\")\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Untuk test Logistic Regression, kelihatan nilai accuracy, precision, recall dan f1-score untuk Test dan Train set itu cukup tepat disekitar 0.8. Untuk values yang \"default\" atau \"1\", test tidak terlalu memiliki performance yang baik dengan skor yang rata-rata rendah karena sample size lebih kecil, oleh karena itu training untuk data yang lulus bayar lebih baik daripada yang gagal bayar. " ] }, { "cell_type": "code", "execution_count": 392, "metadata": {}, "outputs": [], "source": [ "# Model Prediction\n", "\n", "y_pred_train_knn = knn.predict(X_train_scaled)\n", "y_pred_test_knn = knn.predict(X_test_scaled)" ] }, { "cell_type": "code", "execution_count": 393, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " precision recall f1-score support\n", "\n", " 0 0.85 0.96 0.90 1863\n", " 1 0.73 0.36 0.48 509\n", "\n", " accuracy 0.83 2372\n", " macro avg 0.79 0.66 0.69 2372\n", "weighted avg 0.82 0.83 0.81 2372\n", "\n", " precision recall f1-score support\n", "\n", " 0 0.85 0.97 0.90 467\n", " 1 0.74 0.36 0.48 126\n", "\n", " accuracy 0.84 593\n", " macro avg 0.79 0.66 0.69 593\n", "weighted avg 0.82 0.84 0.81 593\n", "\n" ] } ], "source": [ "# Model Evaluation - Train Set & Test Set\n", "\n", "print(classification_report(y_train, y_pred_train_knn))\n", "print(classification_report(y_test, y_pred_test_knn))" ] }, { "cell_type": "code", "execution_count": 394, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Train Set Evaluation (KNN):\n", "accuracy: 0.8339\n", "precision (weighted avg): 0.8209\n", "recall (weighted avg): 0.8339\n", "f1-score (weighted avg): 0.8116\n", "Test Set Evaluation (KNN):\n", "accuracy: 0.8364\n", "precision (weighted avg): 0.8244\n", "recall (weighted avg): 0.8364\n", "f1-score (weighted avg): 0.8133\n" ] } ], "source": [ "# Model Evaluation - Train Set\n", "train_report_knn = classification_report(y_train, y_pred_train_knn, output_dict=True)\n", "train_accuracy_knn = accuracy_score(y_train, y_pred_train_knn)\n", "\n", "print(\"Train Set Evaluation (KNN):\")\n", "#print(classification_report(y_train, y_pred_train_knn))\n", "print(f\"accuracy: {train_accuracy_knn:.4f}\")\n", "print(f\"precision (weighted avg): {train_report_knn['weighted avg']['precision']:.4f}\")\n", "print(f\"recall (weighted avg): {train_report_knn['weighted avg']['recall']:.4f}\")\n", "print(f\"f1-score (weighted avg): {train_report_knn['weighted avg']['f1-score']:.4f}\")\n", "\n", "# Model Evaluation - Test Set\n", "test_report_knn = classification_report(y_test, y_pred_test_knn, output_dict=True)\n", "test_accuracy_knn = accuracy_score(y_test, y_pred_test_knn)\n", "\n", "print(\"Test Set Evaluation (KNN):\")\n", "#print(classification_report(y_test, y_pred_test_knn))\n", "print(f\"accuracy: {test_accuracy_knn:.4f}\")\n", "print(f\"precision (weighted avg): {test_report_knn['weighted avg']['precision']:.4f}\")\n", "print(f\"recall (weighted avg): {test_report_knn['weighted avg']['recall']:.4f}\")\n", "print(f\"f1-score (weighted avg): {test_report_knn['weighted avg']['f1-score']:.4f}\")\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Untuk test KNN, kelihatan nilai accuracy, precision, recall dan f1-score untuk Test dan Train set itu cukup tepat disekitar 0.8. Untuk values yang \"default\" atau \"1\", test tidak terlalu memiliki performance yang baik dengan skor yang rata-rata rendah, sama situasi dengan test sebelumnya. " ] }, { "cell_type": "code", "execution_count": 395, "metadata": {}, "outputs": [], "source": [ "y_pred_train_svc = svc.predict(X_train_scaled)\n", "y_pred_test_svc = svc.predict(X_test_scaled)" ] }, { "cell_type": "code", "execution_count": 396, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " precision recall f1-score support\n", "\n", " 0 0.85 0.97 0.91 1863\n", " 1 0.77 0.38 0.51 509\n", "\n", " accuracy 0.84 2372\n", " macro avg 0.81 0.67 0.71 2372\n", "weighted avg 0.83 0.84 0.82 2372\n", "\n", " precision recall f1-score support\n", "\n", " 0 0.86 0.97 0.91 467\n", " 1 0.77 0.40 0.52 126\n", "\n", " accuracy 0.85 593\n", " macro avg 0.81 0.68 0.72 593\n", "weighted avg 0.84 0.85 0.83 593\n", "\n" ] } ], "source": [ "print(classification_report(y_train, y_pred_train_svc))\n", "print(classification_report(y_test, y_pred_test_svc))" ] }, { "cell_type": "code", "execution_count": 397, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Train Set Evaluation (SVC):\n", "accuracy: 0.8423\n", "precision (weighted avg): 0.8336\n", "recall (weighted avg): 0.8423\n", "f1-score (weighted avg): 0.8204\n", "Test Set Evaluation (SVC):\n", "accuracy: 0.8465\n", "precision (weighted avg): 0.8376\n", "recall (weighted avg): 0.8465\n", "f1-score (weighted avg): 0.8267\n" ] } ], "source": [ "# Model Evaluation - Train Set\n", "train_report_svc = classification_report(y_train, y_pred_train_svc, output_dict=True)\n", "train_accuracy_svc = accuracy_score(y_train, y_pred_train_svc)\n", "\n", "print(\"Train Set Evaluation (SVC):\")\n", "#print(classification_report(y_train, y_pred_train_svc))\n", "print(f\"accuracy: {train_accuracy_svc:.4f}\")\n", "print(f\"precision (weighted avg): {train_report_svc['weighted avg']['precision']:.4f}\")\n", "print(f\"recall (weighted avg): {train_report_svc['weighted avg']['recall']:.4f}\")\n", "print(f\"f1-score (weighted avg): {train_report_svc['weighted avg']['f1-score']:.4f}\")\n", "\n", "# Model Evaluation - Test Set\n", "test_report_svc = classification_report(y_test, y_pred_test_svc, output_dict=True)\n", "test_accuracy_svc = accuracy_score(y_test, y_pred_test_svc)\n", "\n", "print(\"Test Set Evaluation (SVC):\")\n", "#print(classification_report(y_test, y_pred_test_svc))\n", "print(f\"accuracy: {test_accuracy_svc:.4f}\")\n", "print(f\"precision (weighted avg): {test_report_svc['weighted avg']['precision']:.4f}\")\n", "print(f\"recall (weighted avg): {test_report_svc['weighted avg']['recall']:.4f}\")\n", "print(f\"f1-score (weighted avg): {test_report_svc['weighted avg']['f1-score']:.4f}\")\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Untuk test SVC, kelihatan nilai accuracy, precision, recall dan f1-score untuk Test dan Train set itu cukup tepat disekitar 0.8. Untuk values yang \"default\" atau \"1\", test tidak terlalu memiliki performance yang baik dengan skor yang rata-rata rendah, sama situasi dengan test sebelumnya. Oleh Karena SVC memiliki akurasi, precision, recall, dan f-1 score yang paling baik, maka kita akan coba cross-validate dengan SVC dataset. " ] }, { "cell_type": "code", "execution_count": 398, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Non Scaled SVM\n", "Train : 0.00392156862745098\n", "Test : 0.0\n", "\n", "Scaled SVM\n", "Train : 0.5065963060686016\n", "Test : 0.5235602094240838\n" ] } ], "source": [ "# Model Evaluation\n", "\n", "def performance_check(clf, X, y):\n", " y_pred = clf.predict(X)\n", " return f1_score(y, y_pred)\n", "\n", "print('Non Scaled SVM')\n", "print('Train : ', performance_check(svm_non_scaled, X_train, y_train))\n", "print('Test : ', performance_check(svm_non_scaled, X_test, y_test))\n", "print('')\n", "\n", "print('Scaled SVM')\n", "print('Train : ', performance_check(svm_scaled, X_train_scaled, y_train))\n", "print('Test : ', performance_check(svm_scaled, X_test_scaled, y_test))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Bisa diperbandinkan untuk performance chek f-1 score antar train dan test values di SVM. Skor lebih rendah daripada classification report, ini artinya ada masalah fitting atau gunakan hyperparameter tuning nanti. " ] }, { "cell_type": "code", "execution_count": 399, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "SVM - Linear\n", "Train : 0.472258064516129\n", "Test : 0.5025641025641026\n", "\n", "SVM - Polynomial\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Train : 0.5301837270341208\n", "Test : 0.5157894736842106\n", "\n", "SVM - RBF\n", "Train : 0.5065963060686016\n", "Test : 0.5235602094240838\n", "\n" ] } ], "source": [ "# Model Evaluation\n", "\n", "print('SVM - Linear')\n", "print('Train : ', performance_check(svm_linear, X_train_scaled, y_train))\n", "print('Test : ', performance_check(svm_linear, X_test_scaled, y_test))\n", "print('')\n", "\n", "print('SVM - Polynomial')\n", "print('Train : ', performance_check(svm_poly, X_train_scaled, y_train))\n", "print('Test : ', performance_check(svm_poly, X_test_scaled, y_test))\n", "print('')\n", "\n", "print('SVM - RBF')\n", "print('Train : ', performance_check(svm_rbf, X_train_scaled, y_train))\n", "print('Test : ', performance_check(svm_rbf, X_test_scaled, y_test))\n", "print('')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Performance antara Linear, Poly, dan RBF sangat dekat, walaupun yang Polynomial paling baik nilai yang cenderung tinggi untuk Train dan Test. " ] }, { "cell_type": "code", "execution_count": 400, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "SVM - C=0.1\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Train : 0.3203732503888025\n", "Test : 0.3712574850299401\n", "\n", "SVM - C=500\n", "Train : 0.831140350877193\n", "Test : 0.4773662551440329\n", "\n" ] } ], "source": [ "# Model Evaluation\n", "\n", "print('SVM - C=0.1')\n", "print('Train : ', performance_check(svm_rbf_1, X_train_scaled, y_train))\n", "print('Test : ', performance_check(svm_rbf_1, X_test_scaled, y_test))\n", "print('')\n", "\n", "print('SVM - C=500')\n", "print('Train : ', performance_check(svm_rbf_500, X_train_scaled, y_train))\n", "print('Test : ', performance_check(svm_rbf_500, X_test_scaled, y_test))\n", "print('')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Untuk C Regularization test, bisa kelihatan C value yang lebih tinggi artinya akan punya penalty terhadap data yang \"misclassified\" jadi toleransi terhadap data yang salah lebih kecil. Oleh karena itu, SVM yang punya C value yang besar akan cenderung punya akurasi yang baik. " ] }, { "cell_type": "code", "execution_count": 401, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "SVM - gamma=0.1\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Train : 0.5137614678899083\n", "Test : 0.5284974093264249\n", "\n", "SVM - gamma=100\n", "Train : 0.9891411648568608\n", "Test : 0.1566265060240964\n", "\n" ] } ], "source": [ "print('SVM - gamma=0.1')\n", "print('Train : ', performance_check(svm_rbf_500_1, X_train_scaled, y_train))\n", "print('Test : ', performance_check(svm_rbf_500_1, X_test_scaled, y_test))\n", "print('')\n", "\n", "print('SVM - gamma=100')\n", "print('Train : ', performance_check(svm_rbf_500_100, X_train_scaled, y_train))\n", "print('Test : ', performance_check(svm_rbf_500_100, X_test_scaled, y_test))\n", "print('')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Untuk gamma, kelihatan juga value gamma yang lebih besar akan berdampak \"overfitting\" terhadap training dataset. " ] }, { "cell_type": "code", "execution_count": 402, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Decision Tree - Train\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ " precision recall f1-score support\n", "\n", " 0 0.87 0.95 0.91 1863\n", " 1 0.75 0.50 0.60 509\n", "\n", " accuracy 0.86 2372\n", " macro avg 0.81 0.72 0.75 2372\n", "weighted avg 0.85 0.86 0.84 2372\n", "\n", "\n", "Decision Tree - Test\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ " precision recall f1-score support\n", "\n", " 0 0.86 0.95 0.90 467\n", " 1 0.69 0.40 0.51 126\n", "\n", " accuracy 0.83 593\n", " macro avg 0.77 0.68 0.71 593\n", "weighted avg 0.82 0.83 0.82 593\n", "\n" ] } ], "source": [ "# Model Evaluation\n", "\n", "def performance_check(clf, X, y):\n", " y_pred = clf.predict(X)\n", " cm = confusion_matrix(y, y_pred)\n", " disp = ConfusionMatrixDisplay(confusion_matrix=cm)\n", " disp.plot()\n", " plt.show()\n", " print(classification_report(y, y_pred))\n", "\n", "print('Decision Tree - Train')\n", "performance_check(model_dt, X_train_scaled, y_train)\n", "print('')\n", "\n", "print('Decision Tree - Test')\n", "performance_check(model_dt, X_test_scaled, y_test)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita coba buat decision tree untuk melihat rangka model performance seperti TP, TN, FP, dan FN. Kelihatan bahwa data yang True Negative adalah mayoritas untuk Training dan Test Dataset. Ini artinya client yang lunas bayar utang benar diprediksi untuk bayar utang dengan tepat waktu. Di sisi lain, True positive juga cukup signifikan oleh karena banyak orang di train dan test dataset tidak bayar dengan lunas, dan model prediksi dengan benar.\n", "\n", "Untuk masalah yang parah, False Negative adalah situasi jika model tidak prediksi dengan benar jika client tidak bisa bayar utang atau \"default\", tetapi model prediksi bisa lunas bayar. Di train dataset memiliki sekitar 250-an sample yang False Negative, ini sangat tinggi terhadap resiko bank untuk kasih pinjaman atau \"loan\" terhadap client yang tidak mampu bayar.\n", "\n", "Juga, dataset memiliki False Positif yang cukup tinggi juga jika model salah prediksi client yang kelihatannya tidak bisa bayar, tetapi mampu bayar. Ini juga masalah dari model karena client tersebut bisa di \"accuse\" untuk tidak bisa bayar, dan reputasi dari bank bisa terdampak negatif. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Model Evaluation" ] }, { "cell_type": "code", "execution_count": 403, "metadata": {}, "outputs": [], "source": [ "# Calculate predictions for training and test sets\n", "y_train_pred = lg.predict(X_train_final)\n", "y_test_pred = lg.predict(X_test_final)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Untuk Model Evaluasi, Kita akan ambil dua X_train dan X_test file dan akan coba prediksi untuk di-test dengan sample file baru nanti. Kita akan taruh dua file kedalam dua variable baru. " ] }, { "cell_type": "code", "execution_count": 404, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Model Evaluation Metrics:\n", "\n", "Mean Absolute Error (MAE)\n", " - Train Set: 0.17833052276559866\n", " - Test Set: 0.16863406408094436\n", "\n", "\n", "Mean Squared Error (MSE)\n", " - Train Set: 0.17833052276559866\n", " - Test Set: 0.16863406408094436\n", "\n", "\n", "Root Mean Squared Error (RMSE)\n", " - Train Set: 0.4222919875697367\n", " - Test Set: 0.41065078117659093\n", "\n", "\n", "R^2 Score\n", " - Train Set: -0.058094397464005354\n", " - Test Set: -0.007783555963427391\n" ] } ], "source": [ "# Evaluate using metrics MAE, MSE, and Rsquared\n", "mae_train = mean_absolute_error(y_train, y_train_pred)\n", "mae_test = mean_absolute_error(y_test, y_test_pred)\n", "\n", "mse_train = mean_squared_error(y_train, y_train_pred)\n", "mse_test = mean_squared_error(y_test, y_test_pred)\n", "\n", "rmse_train = mean_squared_error(y_train, y_train_pred, squared=False)\n", "rmse_test = mean_squared_error(y_test, y_test_pred, squared=False)\n", "\n", "r2_train = r2_score(y_train, y_train_pred)\n", "r2_test = r2_score(y_test, y_test_pred)\n", "\n", "# Print the evaluation report\n", "print(\"Model Evaluation Metrics:\\n\")\n", "print(f\"Mean Absolute Error (MAE)\\n - Train Set: {mae_train}\\n - Test Set: {mae_test}\\n\")\n", "print()\n", "print(f\"Mean Squared Error (MSE)\\n - Train Set: {mse_train}\\n - Test Set: {mse_test}\\n\")\n", "print()\n", "print(f\"Root Mean Squared Error (RMSE)\\n - Train Set: {rmse_train}\\n - Test Set: {rmse_test}\\n\")\n", "print()\n", "print(f\"R^2 Score\\n - Train Set: {r2_train}\\n - Test Set: {r2_test}\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Di test ini, kita akan nge-test beberapa metrik yang akan mendapatkan insight terhadap dataset Y_train dengan Y_train_pred (akan diprediksi):\n", "\n", "-Mean Absolute Error memiliki akurasi error sebesar 0.17 untuk Train dan Test set. Value lebih rendah lebih akurat. \n", "-Mean Squared Error memiliki akurasi error sebesar 0.17 untuk Train dan Test set. Value lebih rendah lebih akurat. \n", "-Root Mean Squared Error memiliki akurasi error sebesar 0.42 untuk Train dan Test set. Value lebih rendah lebih akurat. \n", "\n", "-R^2 adalah proporsi dari variance target yang bisa dijelaskan oleh karena model tersebut. Disini, value tersebut adalah negatif yang di-train dan test set, artinya model tidak bisa menjelaskan variasi dari target variable \"default_payment_next_month\".\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Cross Validation and Parameter Tuning" ] }, { "cell_type": "code", "execution_count": 405, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
RandomForestClassifier()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "RandomForestClassifier()" ] }, "execution_count": 405, "metadata": {}, "output_type": "execute_result" } ], "source": [ "rf = RandomForestClassifier()\n", "rf.fit(X_train_scaled, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan coba menggunakan Random Forest classifier sebagai baseline model." ] }, { "cell_type": "code", "execution_count": 406, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "F1 Score - Train Set : 0.9911330049261083 \n", "\n", "Classification Report : \n", " precision recall f1-score support\n", "\n", " 0 1.00 1.00 1.00 1863\n", " 1 0.99 0.99 0.99 509\n", "\n", " accuracy 1.00 2372\n", " macro avg 1.00 0.99 0.99 2372\n", "weighted avg 1.00 1.00 1.00 2372\n", " \n", "\n", "Confusion Matrix : \n", " \n" ] }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgMAAAGwCAYAAAA0bWYRAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy80BEi2AAAACXBIWXMAAA9hAAAPYQGoP6dpAAA9CklEQVR4nO3df3zN9f//8fs542zGNj+3s2XmV4b8KtXau3fiTRv5ije9y68spF/0g5S8C0PhzTtJb2/6hX4QelcqectIpiyFhqS9IxrZpsKOzdt+nu8f3jufTpycs3O2w3ndrl1el8ter9fz9TqPl1ycxx6P5+v1MtntdrsAAIBhmf0dAAAA8C+SAQAADI5kAAAAgyMZAADA4EgGAAAwOJIBAAAMjmQAAACDq+HvALxRXl6uo0ePKiwsTCaTyd/hAAA8ZLfbderUKcXExMhsrrrfT8+cOaPi4mKvz2OxWBQSEuKDiC4ul3QycPToUcXGxvo7DACAlw4fPqzGjRtXybnPnDmjBrVCdVreP2PParXq4MGDAZcQXNLJQFhYmCRpiEJlEZUBBKZ5Ofv8HQJQZWynTim21RWOf8+rQnFxsU7LriGq7dV3RbHsWpabq+LiYpKBi0lFa8AiE8kAAlZ4eLi/QwCqXHW0ekO8/K4I5El2l3QyAACAu8wyyexF0mEO4Df5kAwAAAzBLO9+uw/kykAgXxsAAHADlQEAgCGYTJLZi6kJJkk+uCHhokQyAAAwBNoErgXytQEAADdQGQAAGILZ5OXdBBJtAgAALmW0CVwL5GsDAABuoDIAADAEs5d3EwTyb88kAwAAQ6BN4FogXxsAAHADlQEAgCGYTCavXogUyK/DIxkAABgCbQLXSAYAAIbABELXAvnaAADwm/T0dPXp00cxMTEymUxavXq10/6KtsVvlzlz5jjGNG3a9Jz9s2bNcjrP7t27dcMNNygkJESxsbGaPXu2x7FSGQAAGIJJ3v0G7GlRobCwUB07dtSIESPUv3//c/bn5OQ4rf/73//WyJEjNWDAAKft06ZN06hRoxzrYWFhjp9tNpuSkpLUo0cPLVq0SHv27NGIESNUt25d3X333W7HSjIAADAEnzyO2AO9evVSr169XO63Wq1O6++99566deum5s2bO20PCws7Z2yFZcuWqbi4WIsXL5bFYtEVV1yhzMxMzZ0716NkgDYBAAAesNlsTktRUZHX58zLy9OHH36okSNHnrNv1qxZatCgga688krNmTNHpaWljn0ZGRnq0qWLLBaLY1tycrKysrJ04sQJtz+fygAAwBB8dTdBbGys0/YpU6YoNTXVizNLr776qsLCws5pJzz44IO66qqrVL9+fW3dulUTJ05UTk6O5s6dK0nKzc1Vs2bNnI6Jiopy7KtXr55bn08yAAAwBF/dTXD48GGFh4c7tgcHB3sXmKTFixdryJAhCgkJcdo+btw4x88dOnSQxWLRPffco5kzZ/rkcyuQDAAA4IHw8HCnZMBbW7ZsUVZWllauXHnBsQkJCSotLdWhQ4cUHx8vq9WqvLw8pzEV667mGZwPcwYAAIZg9sFSFV555RV17txZHTt2vODYzMxMmc1mRUZGSpISExOVnp6ukpISx5i0tDTFx8e73SKQSAYAAAZhlsnrxRMFBQXKzMxUZmamJOngwYPKzMxUdna2Y4zNZtNbb72lu+6665zjMzIyNG/ePO3atUvff/+9li1bprFjx2ro0KGOL/rBgwfLYrFo5MiR2rt3r1auXKnnnnvOqb3gDtoEAABUge3bt6tbt26O9Yov6JSUFC1dulSStGLFCtntdg0aNOic44ODg7VixQqlpqaqqKhIzZo109ixY52+6CMiIrR+/XqNHj1anTt3VsOGDTV58mSPbiuUJJPdbrdX4hovCjabTRERERqu2rIE9CskYGSLCo/4OwSgythsNkVEN1F+fr5P+/DnfEZEhJ4KqacQL54zcMZu15NnTlRprP5CZQAAYAi8qMg1kgEAgCHwoiLXAvnaAACAG6gMAAAM4eyLiipfGjDpkp1id0EkAwAAQ6BN4FogXxsAAHADlQEAgCFwN4FrJAMAAEOgTeBaIF8bAABwA5UBAIAhVOb9Ar89PlCRDAAADIE2gWuBfG0AAMANVAYAAIZg+t/izfGBimQAAGAItAlcIxkAABgCEwhdC+REBwAAuIHKAADAEGgTuEYyAAAwhLNvLfTu+EAVyIkOAABwA5UBAIAhcGuhayQDAABDMJtMMpu4m+B8aBMAAGBwVAYAAIZAm8A1kgEAgCGQDLhGmwAAAIOjMgAAMAQqA66RDAAADMFkMsnkxd0EpgBOB0gGAACGQGXANeYMAABgcFQGAACGYJZ3vwEH8m/PJAMAAEMwmc4ulT7ed6FcdAI50QEAAG6gMgAAMATT//7z5vhARTIAADAE7iZwjTYBAAAGR2UAAGAIVAZcIxkAABiCWZLZi290s91noVx0aBMAAGBwVAYAAIbA3QSuURkAABiGyYvFU+np6erTp49iYmJkMpm0evVqp/133nmn4+VJFUvPnj2dxhw/flxDhgxReHi46tatq5EjR6qgoMBpzO7du3XDDTcoJCREsbGxmj17tsexkgwAAAyh4gmE3iyeKCwsVMeOHbVgwQKXY3r27KmcnBzH8uabbzrtHzJkiPbu3au0tDStWbNG6enpuvvuux37bTabkpKSFBcXpx07dmjOnDlKTU3Viy++6FGstAkAAKgCvXr1Uq9evX53THBwsKxW63n37du3T+vWrdOXX36pq6++WpL0/PPP6+abb9bf//53xcTEaNmyZSouLtbixYtlsVh0xRVXKDMzU3PnznVKGi6EygAAwBC8aRH8ulVgs9mclqKiokrH9MknnygyMlLx8fG677779Msvvzj2ZWRkqG7duo5EQJJ69Oghs9msbdu2OcZ06dJFFovFMSY5OVlZWVk6ceKE23GQDAAADMEsk9eLJMXGxioiIsKxzJw5s1Lx9OzZU6+99po2btyov/3tb9q8ebN69eqlsrIySVJubq4iIyOdjqlRo4bq16+v3Nxcx5ioqCinMRXrFWPcQZsAAAAPHD58WOHh4Y714ODgSp1n4MCBjp/bt2+vDh06qEWLFvrkk0/UvXt3r+P0BJUBAIAh+KpNEB4e7rRUNhn4rebNm6thw4bav3+/JMlqterYsWNOY0pLS3X8+HHHPAOr1aq8vDynMRXrruYinA/JAADAEKr7bgJPHTlyRL/88ouio6MlSYmJiTp58qR27NjhGPPxxx+rvLxcCQkJjjHp6ekqKSlxjElLS1N8fLzq1avn9meTDAAAUAUKCgqUmZmpzMxMSdLBgweVmZmp7OxsFRQU6NFHH9Xnn3+uQ4cOaePGjerbt69atmyp5ORkSVKbNm3Us2dPjRo1Sl988YU+++wzjRkzRgMHDlRMTIwkafDgwbJYLBo5cqT27t2rlStX6rnnntO4ceM8ipU5AwAAQ6juFxVt375d3bp1c6xXfEGnpKRo4cKF2r17t1599VWdPHlSMTExSkpK0vTp053aDsuWLdOYMWPUvXt3mc1mDRgwQPPnz3fsj4iI0Pr16zV69Gh17txZDRs21OTJkz26rVCSTHa7/ZJ99YLNZlNERISGq7YsAfyYSBjbosIj/g4BqDI2m00R0U2Un5/vNCnP558REaE1DaJV21z5gnhhebn+3y85VRqrv9AmAADA4GgTAAAMwWzy8hXGAVyAJhkAABhCdc8ZuJSQDAAADIFkwDXmDAAAYHBUBgAAhmD633/eHB+oSAYAAIbg7VMEq/oJhP5EmwAAAIOjMmAwLa9PUNLD96rJle1VN9qqhbeP1K41Hzn2B9cO1Z+n/VUd+ySrdv16+uVQtj5euFhbXnnD6TzNrr1KfVMnqNnVV6q8rExHdu/V/L5DVXLmjCQptF5dDXxmutr36iF7ebm+eu/fWvXoZBUVnq7W6wXcsfml15T+0uv6JfvsA56i27RS78cfVrvkbhc4EpcSs7z7DTiQf3smGTCY4NqhOrLnG219baXuXfHyOftvnTVF8TderyUjH9QvPxxWm+43atC8p5Wfk6fda9MknU0EHlz9htY9s0ArH5mk8tJSNW7fVvbycsd5Rix+XhHWSD3XZ7CCatZQyqK5GvKP2Vo8fEy1XSvgrnqXRavftImKbNlMstuVsewtLbx9pJ7Y+m/FtI33d3jwEe4mcO2iSHQWLFigpk2bKiQkRAkJCfriiy/8HVLA2rt+k96fNkeZH6w77/7m13XW58ve0n+2ZOiX7CP6dMkyHdnzjZpe3ckx5i9/S9XHCxfro2cWKGfff5T33ffa8c4alRYXS5Ks8S3VLqmbXr//UR3a/pUOZHypFeMn6epbb1GENao6LhPwSIebb1L7nn9SVMtmirq8ufqlTlBwnVAd/PIrf4cGVAu/JwMrV67UuHHjNGXKFO3cuVMdO3ZUcnLyOe9wRvX4/vMd6tD7JtWNPvse7FZd/qCols31zcZ0SVJYowZqfu1VOvXTL3p042rNPviVxq37l1okXuM4R/OEzio8cVLZX+12bPv24y2yl5er2TVXVu8FAR4qLyvTl2+9p+LC/6rZtVf5Oxz4kskkkxdLIM8g9HsyMHfuXI0aNUrDhw9X27ZttWjRIoWGhmrx4sX+Ds2QVj4ySTnffqdZ+7drwcmDemD163pz3BPa/9k2SVLDpnGSpP/313H6dOlyPd9vqA7v2qOHP1yhyBbNJEnhkY106qdfnM5bXlamwhMnFR7VqHovCHDTj1/v00OR8RpTr4WWP/RX3fPmS4pp08rfYcGHTD5YApVf5wwUFxdrx44dmjhxomOb2WxWjx49lJGRcc74oqIiFRUVOdZtNlu1xGkk3e4brmbXXKUFt96p44d/1OXXJ2jQ3LNzBr7d9KlM/3s495bFbyjj9VWSpMO79iq+6x/1h2G3a/WUWf4MH6i0qFYt9ETGOv3Xdko7312rV+8Zq3Hr3iIhgCH4NRn4+eefVVZWpqgo5z5yVFSUvv3223PGz5w5U1OnTq2u8AynZkiI+qZO0KKBd+nrjz6WdPa3pcYdrtBND92rbzd9qvzcs+2bnG+/czo299vvVD/2MkmS7dhPCmvUwGm/OShItevVlS3vp2q4EsBzNSwWR3Ur7soO+mHHLm3652INeZ4EN1AwgdA1v7cJPDFx4kTl5+c7lsOHD/s7pIASVLOGalgsstvtTtvLy8ocFYFffjisk0dzFXV5c6cxkZc3d9yW9f22Hapdr66adGrv2B/f9XqZzGYmZOGSYS8vV8mvKpG49HkzX8AxbyBA+bUy0LBhQwUFBSkvL89pe15enqxW6znjg4ODFRwcXF3hBaTg2qFq1KKpY71h01g17tBWhcdP6sSRo/pPeob6P/2ESv57Rr9kH1GrG67TdYNv1b8e/7+KzPp5C9XniUf04559Orx7r64bcqusrVrqxSH3SJJys/br6/WbNHTBbC1/cKKCatbQwGee0vZ/va/83LzfhgT43buTZ6ldUlfVi71MRacK9MWq9/SfLRl64L03LnwwLhm8wtg1vyYDFotFnTt31saNG9WvXz9JUnl5uTZu3KgxY7gfvSrEXdVR49a95Vj/y99SJUkZb6zSq/eM08t33q9+Ux/XiMXPK7ReXR3PPqL3pv5N6S+/7jjm4wWvqGZIiG792xTVrldXR/Z8o+f6DNLPB39wjFk84gENnPuUHv5whezl5dr53lqtGj+52q4T8MSpn37WklFjZcs9plrhYbqsXRs98N4batu9i79DA6qFyf7bmnA1W7lypVJSUvTCCy/o2muv1bx587Rq1Sp9++2358wl+C2bzaaIiAgNV21ZArqbAyNbVHjE3yEAVcZmsykiuony8/MVHh5edZ8REaEtMbGqY658d7ygvFw3HD1cpbH6i9+fQHj77bfrp59+0uTJk5Wbm6tOnTpp3bp1F0wEAADwBC8qcs3vyYAkjRkzhrYAAAB+clEkAwAAVDUqA66RDAAADMHb2wMD+dbCS+o5AwAAwPeoDAAADIE2gWskAwAAQ6BN4BptAgAADI7KAADAEGgTuEYyAAAwBLPJJLMX3+jeHHuxIxkAABgClQHXmDMAAIDBURkAABiCSV7eTRDAL8QjGQAAGILJfHap9PF+fcdv1aJNAACAwVEZAAAYg5cPHQrkGYQkAwAAQ+BuAtdoEwAAYHBUBgAAhnC2MuDNuwl8GMxFhmQAAGAItAlco00AAEAVSE9PV58+fRQTEyOTyaTVq1c79pWUlGjChAlq3769ateurZiYGA0bNkxHjx51OkfTpk0db1usWGbNmuU0Zvfu3brhhhsUEhKi2NhYzZ492+NYSQYAAIZQ8W4CbxZPFBYWqmPHjlqwYME5+06fPq2dO3dq0qRJ2rlzp9555x1lZWXplltuOWfstGnTlJOT41geeOABxz6bzaakpCTFxcVpx44dmjNnjlJTU/Xiiy96FCttAgCAIVR3m6BXr17q1avXefdFREQoLS3Nads//vEPXXvttcrOzlaTJk0c28PCwmS1Ws97nmXLlqm4uFiLFy+WxWLRFVdcoczMTM2dO1d3332327FSGQAAGMJvy+2VWaSzv43/eikqKvJJfPn5+TKZTKpbt67T9lmzZqlBgwa68sorNWfOHJWWljr2ZWRkqEuXLrJYLI5tycnJysrK0okTJ9z+bJIBAAA8EBsbq4iICMcyc+ZMr8955swZTZgwQYMGDVJ4eLhj+4MPPqgVK1Zo06ZNuueeezRjxgw99thjjv25ubmKiopyOlfFem5urtufT5sAAGAIvmoTHD582OkLOzg42Ku4SkpKdNttt8lut2vhwoVO+8aNG+f4uUOHDrJYLLrnnns0c+ZMrz/310gGAACG4KtkIDw83CkZ8EZFIvDDDz/o448/vuB5ExISVFpaqkOHDik+Pl5Wq1V5eXlOYyrWXc0zOB/aBAAA+EFFIvDdd99pw4YNatCgwQWPyczMlNlsVmRkpCQpMTFR6enpKikpcYxJS0tTfHy86tWr53YsVAYAAIZgMptkMnvxBEK7Z8cWFBRo//79jvWDBw8qMzNT9evXV3R0tG699Vbt3LlTa9asUVlZmaPHX79+fVksFmVkZGjbtm3q1q2bwsLClJGRobFjx2ro0KGOL/rBgwdr6tSpGjlypCZMmKCvv/5azz33nJ599lmPYiUZAAAYQnXfWrh9+3Z169bNsV7R/09JSVFqaqref/99SVKnTp2cjtu0aZO6du2q4OBgrVixQqmpqSoqKlKzZs00duxYp3kEERERWr9+vUaPHq3OnTurYcOGmjx5ske3FUokAwAAVImuXbvKbre73P97+yTpqquu0ueff37Bz+nQoYO2bNnicXy/RjIAADCEyjxF8LfHByqSAQCAIfCiIte4mwAAAIOjMgAAMIRfP1K4sscHKpIBAIAhmORlm8BnkVx8SAYAAIZAZcA15gwAAGBwVAYAAMbg5d0EgdwnIBkAABgCbQLXaBMAAGBwVAYAAIZgMp9dvDk+UJEMAAAMgTaBawGc5wAAAHdQGQAAGIPZdHbx5vgARTIAADAG3lTkEskAAMAQmDPgGnMGAAAwOCoDAABjYM6ASyQDAABjYM6AS7QJAAAwOCoDAABDMJlNMnlR6vfm2IsdyQAAwBhoE7hEmwAAAIOjMgAAMASTycs2QQBXBkgGAADGQJvAJbeSgffff9/tE95yyy2VDgYAAFQ/t5KBfv36uXUyk8mksrIyb+IBAKBqmOXlQ4d8FslFx61koLy8vKrjAACgSvFuAte8mjNw5swZhYSE+CoWAACqDo8jdsnjokdZWZmmT5+uyy67THXq1NH3338vSZo0aZJeeeUVnwcIAACqlsfJwNNPP62lS5dq9uzZslgsju3t2rXTyy+/7NPgAADwmYq7CbxZApTHycBrr72mF198UUOGDFFQUJBje8eOHfXtt9/6NDgAAHzFZPZ+CVQeX9qPP/6oli1bnrO9vLxcJSUlPgkKAABUH4+TgbZt22rLli3nbP/Xv/6lK6+80idBAQDgc7QJXPL4boLJkycrJSVFP/74o8rLy/XOO+8oKytLr732mtasWVMVMQIA4DXeWuiax5WBvn376oMPPtCGDRtUu3ZtTZ48Wfv27dMHH3ygm266qSpiBAAAVahSzxm44YYblJaW5utYAACoOrybwKVKP3Ro+/bt2rdvn6Sz8wg6d+7ss6AAAPA5HjrkksfJwJEjRzRo0CB99tlnqlu3riTp5MmT+sMf/qAVK1aocePGvo4RAABUIY/nDNx1110qKSnRvn37dPz4cR0/flz79u1TeXm57rrrrqqIEQAAr1W8m8CbxRPp6enq06ePYmJiZDKZtHr1aqf9drtdkydPVnR0tGrVqqUePXrou+++cxpz/PhxDRkyROHh4apbt65GjhypgoICpzG7d+/WDTfcoJCQEMXGxmr27Nke/9l4nAxs3rxZCxcuVHx8vGNbfHy8nn/+eaWnp3scAAAA1aKiTeDN4oHCwkJ17NhRCxYsOO/+2bNna/78+Vq0aJG2bdum2rVrKzk5WWfOnHGMGTJkiPbu3au0tDStWbNG6enpuvvuux37bTabkpKSFBcXpx07dmjOnDlKTU3Viy++6FGsHrcJYmNjz/twobKyMsXExHh6OgAAqom3zwo4e6zNZnPaGhwcrODg4HNG9+rVS7169Trvmex2u+bNm6cnn3xSffv2lXT2Cb9RUVFavXq1Bg4cqH379mndunX68ssvdfXVV0uSnn/+ed188836+9//rpiYGC1btkzFxcVavHixLBaLrrjiCmVmZmru3LlOScOFeFwZmDNnjh544AFt377dsW379u166KGH9Pe//93T0wEAcEmJjY1VRESEY5k5c6bH5zh48KByc3PVo0cPx7aIiAglJCQoIyNDkpSRkaG6des6EgFJ6tGjh8xms7Zt2+YY06VLF6d3BSUnJysrK0snTpxwOx63KgP16tVz6pUUFhYqISFBNWqcPby0tFQ1atTQiBEj1K9fP7c/HACA6lKZvv9vj5ekw4cPKzw83LH9fFWBC8nNzZUkRUVFOW2Piopy7MvNzVVkZKTT/ho1aqh+/fpOY5o1a3bOOSr21atXz6143EoG5s2b59bJAAC4aPno1sLw8HCnZCAQuJUMpKSkVHUcAAAYhtVqlSTl5eUpOjrasT0vL0+dOnVyjDl27JjTcaWlpTp+/LjjeKvVqry8PKcxFesVY9zh1QsZz5w5I5vN5rQAAHAxqu5bC39Ps2bNZLVatXHjRsc2m82mbdu2KTExUZKUmJiokydPaseOHY4xH3/8scrLy5WQkOAYk56e7jSxPy0tTfHx8W63CKRKJAOFhYUaM2aMIiMjVbt2bdWrV89pAQDgolTNtxYWFBQoMzNTmZmZks5OGszMzFR2drZMJpMefvhhPfXUU3r//fe1Z88eDRs2TDExMY65d23atFHPnj01atQoffHFF/rss880ZswYDRw40HH33uDBg2WxWDRy5Ejt3btXK1eu1HPPPadx48Z5FKvHtxY+9thj2rRpkxYuXKg77rhDCxYs0I8//qgXXnhBs2bN8vR0AAAEpO3bt6tbt26O9Yov6JSUFC1dulSPPfaYCgsLdffdd+vkyZP64x//qHXr1ikkJMRxzLJlyzRmzBh1795dZrNZAwYM0Pz58x37IyIitH79eo0ePVqdO3dWw4YNNXnyZI9uK5Qkk91ut3tyQJMmTfTaa6+pa9euCg8P186dO9WyZUu9/vrrevPNN7V27VqPAvCGzWZTRESEhqu2LArcZ0bD2BYVHvF3CECVsdlsiohuovz8/CqblFfxXfHTHX9SuKXSr+SRrbhUjV7/uEpj9ReP2wTHjx9X8+bNJZ2dUXn8+HFJ0h//+EeeQAgAuGiZzCavl0DlcTLQvHlzHTx4UJLUunVrrVq1SpL0wQcfOF5cBAAALh0eJwPDhw/Xrl27JEmPP/64FixYoJCQEI0dO1aPPvqozwMEAMAnTCbvlwDlcfNk7Nixjp979Oihb7/9Vjt27FDLli3VoUMHnwYHAIDPmOXlQ4d8FslFp/IzKf4nLi5OcXFxvogFAIAq46vHEQcit5KBX9/GcCEPPvhgpYMBAADVz61k4Nlnn3XrZCaTyS/JwLycfQF3mwdQoeyTVf4OAagyZaf/W30f5qN3EwQit5KBirsHAAC4ZHk7CTCA2wQBPB0CAAC4w+sJhAAAXBKoDLhEMgAAMAhvnxUQuMkAbQIAAAyOygAAwBjM5rOLN8cHqEpd2ZYtWzR06FAlJibqxx9/lCS9/vrr+vTTT30aHAAAPsPjiF3yOBl4++23lZycrFq1aumrr75SUVGRJCk/P18zZszweYAAAKBqeZwMPPXUU1q0aJFeeukl1axZ07H9+uuv186dO30aHAAAPkNlwCWP5wxkZWWpS5cu52yPiIjQyZMnfRETAAC+x62FLnlcGbBardq/f/852z/99FM1b97cJ0EBAOBzFRMIvVkClMdXNmrUKD300EPatm2bTCaTjh49qmXLlmn8+PG67777qiJGAABQhTxuEzz++OMqLy9X9+7ddfr0aXXp0kXBwcEaP368HnjggaqIEQAA79EmcMnjZMBkMumJJ57Qo48+qv3796ugoEBt27ZVnTp1qiI+AAB8g2TApUo/dMhisaht27a+jAUAAPiBx8lAt27dZPqd7Ojjjz/2KiAAAKoElQGXPE4GOnXq5LReUlKizMxMff3110pJSfFVXAAA+BaPI3bJ42Tg2WefPe/21NRUFRQUeB0QAACoXj5Lc4YOHarFixf76nQAAPgWTyB0yWdvLczIyFBISIivTgcAgG+Z5OWcAZ9FctHxOBno37+/07rdbldOTo62b9+uSZMm+SwwAABQPTxOBiIiIpzWzWaz4uPjNW3aNCUlJfksMAAAfIq7CVzyKBkoKyvT8OHD1b59e9WrV6+qYgIAwOdMZrNMXtwR4M2xFzuPriwoKEhJSUm8nRAAcAnydvJg4FYGPE5z2rVrp++//74qYgEAAH7gcTLw1FNPafz48VqzZo1ycnJks9mcFgAALkrcWuiS23MGpk2bpkceeUQ333yzJOmWW25xeiyx3W6XyWRSWVmZ76MEAMBbTCB0ye1kYOrUqbr33nu1adOmqowHAABUM7eTAbvdLkm68cYbqywYAACqDO8mcMmjWwt/722FAABc1GgTuORRMtCqVasLJgTHjx/3KiAAAFC9PEoGpk6des4TCAEAuCRUc2WgadOm+uGHH87Zfv/992vBggXq2rWrNm/e7LTvnnvu0aJFixzr2dnZuu+++7Rp0ybVqVNHKSkpmjlzpmrU8NmrhSR5mAwMHDhQkZGRPg0AAIBqUc3JwJdfful0h93XX3+tm266SX/5y18c20aNGqVp06Y51kNDQx0/l5WVqXfv3rJardq6datycnI0bNgw1axZUzNmzKj8dZyH28kA8wUAAHBfo0aNnNZnzZqlFi1aOE3EDw0NldVqPe/x69ev1zfffKMNGzYoKipKnTp10vTp0zVhwgSlpqbKYrH4LFa3p0ZW3E0AAMAlqeJuAm8W6ZyH7RUVFV3wo4uLi/XGG29oxIgRTr9cL1u2TA0bNlS7du00ceJEnT592rEvIyND7du3V1RUlGNbcnKybDab9u7d68M/GA8qA+Xl5T79YAAAqpWP2gSxsbFOm6dMmaLU1NTfPXT16tU6efKk7rzzTse2wYMHKy4uTjExMdq9e7cmTJigrKwsvfPOO5Kk3Nxcp0RAkmM9Nze38tdxHr6dgQAAwMXKR8nA4cOHFR4e7tgcHBx8wUNfeeUV9erVSzExMY5td999t+Pn9u3bKzo6Wt27d9eBAwfUokWLysdZCYH7BAUAAKpAeHi403KhZOCHH37Qhg0bdNddd/3uuISEBEnS/v37JUlWq1V5eXlOYyrWXc0zqCySAQCAMfhozoCnlixZosjISPXu3ft3x2VmZkqSoqOjJUmJiYnas2ePjh075hiTlpam8PBwtW3btlKxuEKbAABgDCZ52Sbw/JDy8nItWbJEKSkpTs8GOHDggJYvX66bb75ZDRo00O7duzV27Fh16dJFHTp0kCQlJSWpbdu2uuOOOzR79mzl5ubqySef1OjRo91qTXiCZAAAgCqyYcMGZWdna8SIEU7bLRaLNmzYoHnz5qmwsFCxsbEaMGCAnnzySceYoKAgrVmzRvfdd58SExNVu3ZtpaSkOD2XwFdIBgAAxuCHdxMkJSWd99b82NjYc54+eD5xcXFau3atx5/rKZIBAIAx8KIil5hACACAwVEZAAAYg6nydwQ4jg9QJAMAAGOgTeBS4KY5AADALVQGAADGQGXAJZIBAIAxmMze9f2ZMwAAwCXObDq7eHN8gArcNAcAALiFygAAwBhoE7hEMgAAMAYmELoUuGkOAABwC5UBAIAxmL18AqE3x17kSAYAAMZAm8ClwE1zAACAW6gMAACMgbsJXCIZAAAYg0letgl8FslFJ3DTHAAA4BYqAwAAY+BuApdIBgAAxsDdBC6RDAAAjIEJhC4F7pUBAAC3UBkAABiDyctXGNMmAADgEkebwKXAvTIAAOAWKgMAAGPgbgKXSAYAAMZAm8ClwL0yAADgFioDAABjMHt5N4E3x17kSAYAAMbAnAGXaBMAAGBwVAYAAMbABEKXSAYAAMbAnAGXSAYAAMZgMnlZGQjcZCBwax4AAMAtVAYAAMbA3QQukQwAAIyBCYQuBe6VAQAAt1AZAAAYA3cTuERlAABgDBVtAm8WD6SmpspkMjktrVu3duw/c+aMRo8erQYNGqhOnToaMGCA8vLynM6RnZ2t3r17KzQ0VJGRkXr00UdVWlrqkz+OX6MyAABAFbniiiu0YcMGx3qNGv/3tTt27Fh9+OGHeuuttxQREaExY8aof//++uyzzyRJZWVl6t27t6xWq7Zu3aqcnBwNGzZMNWvW1IwZM3waJ8kAAMAY/HA3QY0aNWS1Ws/Znp+fr1deeUXLly/Xn/70J0nSkiVL1KZNG33++ee67rrrtH79en3zzTfasGGDoqKi1KlTJ02fPl0TJkxQamqqLBZL5a/lN2gTAACMwWz2fpFks9mclqKiIpcf+d133ykmJkbNmzfXkCFDlJ2dLUnasWOHSkpK1KNHD8fY1q1bq0mTJsrIyJAkZWRkqH379oqKinKMSU5Ols1m0969e337R+PTswEAEOBiY2MVERHhWGbOnHnecQkJCVq6dKnWrVunhQsX6uDBg7rhhht06tQp5ebmymKxqG7duk7HREVFKTc3V5KUm5vrlAhU7K/Y50u0CeCWE0dz9O6TM7U3bZOKT/9XjZo3VcoLzyjuqo7+Dg34XR8u+0hrl6932hbVuJEmv/C4JKmkuETvvPy+dqRnqqSkVG2vitft9w9QeL0wSVKBrVBL5yzT0UM5KrQVqk7dOupwXTvdknKzaoWGVPv1wBtetgl09tjDhw8rPDzcsTU4OPi8o3v16uX4uUOHDkpISFBcXJxWrVqlWrVqeRGH75EM4IIKT5zUnO79Fd8lUWPefU1hDRvo2P6DCq0b4e/QALdEx1n1wFP3ONaDgv6vKPqvl97T3i/3aeTEYaoVGqJVi97VS08v1SN/f0CSZDaZ1OG6duozrJfqRNTWT0d/1qqF72jFqdMa/tjQar8WeMFHDx0KDw93SgbcVbduXbVq1Ur79+/XTTfdpOLiYp08edKpOpCXl+eYY2C1WvXFF184naPiboPzzUPwhl/bBOnp6erTp49iYmJkMpm0evVqf4YDF9bPXaj6jaOV8sJcNbv6SjVs2kRte9yoRs2b+js0wC1ms1kR9cMdS52IOpKk/xb+Vxnrv1D/u25RfMfL1eTyWA19+HZ9v++QDn77gyQpNCxUXXr/QXGXx6pBZH217tRKN/S+Xvv3fu/PS0JlVEwg9GbxQkFBgQ4cOKDo6Gh17txZNWvW1MaNGx37s7KylJ2drcTERElSYmKi9uzZo2PHjjnGpKWlKTw8XG3btvUqlt/yazJQWFiojh07asGCBf4MAxewa22amlzZQS8OvVePxnXS04k9tWXJcn+HBbjtp6M/6693TNXkEU9ryZw3dPzYCUlS9v4jKistU+tOrRxjrbFRqteong7uO3Tec538JV+7tu7R5e1aVEfouISNHz9emzdv1qFDh7R161b9+c9/VlBQkAYNGqSIiAiNHDlS48aN06ZNm7Rjxw4NHz5ciYmJuu666yRJSUlJatu2re644w7t2rVLH330kZ588kmNHj3aZWuisvzaJujVq5dTT+VCioqKnGZt2my2qggLv/HzwWylv/yGejxwl3qOH6Mfdu7SqvGTVaNmTSUO/Yu/wwN+V9P4Jrpj7EBFNW6k/OM2rV2+XnMfW6An/zlethOnVKNGkELrOPdvw+vVke3EKadti//2unZv26uSohK1v7athjx0W3VeBnzhV3cEVPp4Dxw5ckSDBg3SL7/8okaNGumPf/yjPv/8czVq1EiS9Oyzz8psNmvAgAEqKipScnKy/vnPfzqODwoK0po1a3TfffcpMTFRtWvXVkpKiqZNm1b5a3DhkpozMHPmTE2dOtXfYRiOvbxccVd1UL+pZydcNenUTke/yVL6K2+QDOCid8XVbRw/X9YsRk3j4zRp+FPauWWXagbXdPs8A0b11c2Dk3Tsx5/03qtr9fZL72vg6AFVETKqSjU/Z2DFihW/uz8kJEQLFiz43ep4XFyc1q5d69HnVsYldWvhxIkTlZ+f71gOHz7s75AMIcIaqejWlztts8a31PHDP/opIqDyQuvUUuRljfRTzs8Krxem0tIynS74r9MY24kCx90EFSLqh8saG6UO17XToDG3asvarco/TnUSgeGSSgaCg4MdszgrO5sTnmtx3dXK++6A07a8775XgyaN/RQRUHln/lukn3N+Vnj9cDVp2VhBNYKUtes7x/68I8d04qcTatamqctz2O12SVJpie+fEY8qZDJ5+W6CwH1R0SXVJoB/dH/gLs3+05/17znPq3P//6dD2zP16ZLlGvL83/wdGnBB77z8vtonXKH6kfWU/0u+Plz2kcxms66+8UrVql1LiUnX6u2X3lftOqEKCQ3WqkXvqlnrODVrHSdJ+vrLfTp18pTiLo9VcK1g5fyQq3cXr1Hztk3VIKq+n68OHvHD44gvFSQDuKCmnTvp3hUvafXkWfpw5nNq2DRWf5mdqoSBf/Z3aMAFnfwlX0tmv3H2gUERddTiimYaP/dBhf3v9sJbR/WV2WTSSzOWqrSkTG2uitft9/d3HG+x1NRn6z7X2y+9p9KSUtVrWFcd/9BeSX/p7q9LAnzOr8lAQUGB9u/f71g/ePCgMjMzVb9+fTVp0sSPkeG3OvTqoQ69elx4IHCRGTHhjt/dX9NSU7ffP0C333/+yYCtOrbU+GcerIrQUN189NChQOTXZGD79u3q1q2bY33cuHGSpJSUFC1dutRPUQEAApLZdHbx5vgA5ddkoGvXro6JOAAAwD+YMwAAMAbaBC6RDAAAjIG7CVwiGQAAGAOVAZcC98oAAIBbqAwAAAzBZDLJ5EWp35tjL3YkAwAAY6BN4FLgXhkAAHALlQEAgDFQGXCJZAAAYAwmL59AGMBzBgI3zQEAAG6hMgAAMAbaBC6RDAAAjIEnELoUuGkOAABwC5UBAIAxmExetgkCtzJAMgAAMAbaBC6RDAAAjIEJhC4F7pUBAAC3UBkAABiD2cuHDnlz7EWOZAAAYAy0CVwK3CsDAABuoTIAADAG7iZwiWQAAGAMtAlcCtwrAwAAbqEyAAAwBtoELpEMAACMgTaBS4F7ZQAAwC1UBgAAxmA2n128OT5AkQwAAAzBZDLJ5EXf35tjL3YkAwAAY+AVxi4Fbs0DAAC4hcoAAMAYuLXQJZIBAIBBeHlrYQAX0wP3ygAA8KOZM2fqmmuuUVhYmCIjI9WvXz9lZWU5jenatatjYmPFcu+99zqNyc7OVu/evRUaGqrIyEg9+uijKi0t9WmsVAYAAMZQzW2CzZs3a/To0brmmmtUWlqqv/71r0pKStI333yj2rVrO8aNGjVK06ZNc6yHhoY6fi4rK1Pv3r1ltVq1detW5eTkaNiwYapZs6ZmzJhR+Wv5DZIBAIAxVPNzBtatW+e0vnTpUkVGRmrHjh3q0qWLY3toaKisVut5z7F+/Xp988032rBhg6KiotSpUydNnz5dEyZMUGpqqiwWi+fXcR60CQAA8IDNZnNaioqK3DouPz9fklS/fn2n7cuWLVPDhg3Vrl07TZw4UadPn3bsy8jIUPv27RUVFeXYlpycLJvNpr179/rgas6iMgAAMAYftQliY2OdNk+ZMkWpqam/e2h5ebkefvhhXX/99WrXrp1j++DBgxUXF6eYmBjt3r1bEyZMUFZWlt555x1JUm5urlMiIMmxnpubW/lr+Q2SAQCAMfjoRUWHDx9WeHi4Y3NwcPAFDx09erS+/vprffrpp07b7777bsfP7du3V3R0tLp3764DBw6oRYsWlY/VQ7QJAADwQHh4uNNyoWRgzJgxWrNmjTZt2qTGjRv/7tiEhARJ0v79+yVJVqtVeXl5TmMq1l3NM6gMkgEAgDFUtAm8WTxgt9s1ZswYvfvuu/r444/VrFmzCx6TmZkpSYqOjpYkJSYmas+ePTp27JhjTFpamsLDw9W2bVuP4vk9tAkAAAZh+t/izfHuGz16tJYvX6733ntPYWFhjh5/RESEatWqpQMHDmj58uW6+eab1aBBA+3evVtjx45Vly5d1KFDB0lSUlKS2rZtqzvuuEOzZ89Wbm6unnzySY0ePdqt9oS7qAwAAIyhmisDCxcuVH5+vrp27aro6GjHsnLlSkmSxWLRhg0blJSUpNatW+uRRx7RgAED9MEHHzjOERQUpDVr1igoKEiJiYkaOnSohg0b5vRcAl+gMgAAQBWw2+2/uz82NlabN2++4Hni4uK0du1aX4V1XiQDAABj4EVFLpEMAAAMonrnDFxKmDMAAIDBURkAABgDbQKXSAYAAMZAl8Al2gQAABgclQEAgEFQGnCFZAAAYAzMGXCJNgEAAAZHZQAAYAwmeVkZ8FkkFx2SAQCAQTBnwBWSAQCAMTBnwCXmDAAAYHBUBgAABkGbwBWSAQCAMdAmcIk2AQAABkdlAABgDFQGXCIZAAAYBHMGXKFNAACAwVEZAAAYgslkksmLUr83x17sSAYAAMbAnAGXaBMAAGBwVAYAAAbBBEJXSAYAAAbhZZuAZAAAgEsccwZcYs4AAAAGR2UAAGAQzBlwhWQAAGAMtAlcok0AAIDBURkAABgDXQKXSAYAAAZBNuAKbQIAAAyOygAAwBiYQOgSyQAAwBhIBlyiTQAAgMFRGQAAGAQTCF0hGQAAGINJXrYJfBbJRYdkAABgDMwZcIk5AwAAGByVAQCAQTBnwBWSAQCAMdAmcOmSTgbsdrskyXbqlJ8jAapO2en/+jsEoMrYTp+R9H//nlfpZ3n5XRHI3zWXdDJw6n//Y2JbXeHnSAAA3jh16pQiIiKq5NwWi0VWq9Un3xVWq1UWi8UHUV1cTPbqSMeqSHl5uY4ePaqwsDCZArh8czGx2WyKjY3V4cOHFR4e7u9wAJ/i73f1s9vtOnXqlGJiYmQ2V92c9jNnzqi4uNjr81gsFoWEhPggoovLJV0ZMJvNaty4sb/DMKTw8HD+sUTA4u939aqqisCvhYSEBOSXuK9wayEAAAZHMgAAgMGRDMAjwcHBmjJlioKDg/0dCuBz/P2GUV3SEwgBAID3qAwAAGBwJAMAABgcyQAAAAZHMgAAgMGRDMBtCxYsUNOmTRUSEqKEhAR98cUX/g4J8In09HT16dNHMTExMplMWr16tb9DAqoVyQDcsnLlSo0bN05TpkzRzp071bFjRyUnJ+vYsWP+Dg3wWmFhoTp27KgFCxb4OxTAL7i1EG5JSEjQNddco3/84x+Szr4XIjY2Vg888IAef/xxP0cH+I7JZNK7776rfv36+TsUoNpQGcAFFRcXa8eOHerRo4djm9lsVo8ePZSRkeHHyAAAvkAygAv6+eefVVZWpqioKKftUVFRys3N9VNUAABfIRkAAMDgSAZwQQ0bNlRQUJDy8vKctufl5clqtfopKgCAr5AM4IIsFos6d+6sjRs3OraVl5dr48aNSkxM9GNkAABfqOHvAHBpGDdunFJSUnT11Vfr2muv1bx581RYWKjhw4f7OzTAawUFBdq/f79j/eDBg8rMzFT9+vXVpEkTP0YGVA9uLYTb/vGPf2jOnDnKzc1Vp06dNH/+fCUkJPg7LMBrn3zyibp163bO9pSUFC1durT6AwKqGckAAAAGx5wBAAAMjmQAAACDIxkAAMDgSAYAADA4kgEAAAyOZAAAAIMjGQAAwOBIBgAAMDiSAcBLd955p/r16+dY79q1qx5++OFqj+OTTz6RyWTSyZMnXY4xmUxavXq12+dMTU1Vp06dvIrr0KFDMplMyszM9Oo8AKoOyQAC0p133imTySSTySSLxaKWLVtq2rRpKi0trfLPfueddzR9+nS3xrrzBQ4AVY0XFSFg9ezZU0uWLFFRUZHWrl2r0aNHq2bNmpo4ceI5Y4uLi2WxWHzyufXr1/fJeQCgulAZQMAKDg6W1WpVXFyc7rvvPvXo0UPvv/++pP8r7T/99NOKiYlRfHy8JOnw4cO67bbbVLduXdWvX199+/bVoUOHHOcsKyvTuHHjVLduXTVo0ECPPfaYfvt6j9+2CYqKijRhwgTFxsYqODhYLVu21CuvvKJDhw45Xo5Tr149mUwm3XnnnZLOviJ65syZatasmWrVqqWOHTvqX//6l9PnrF27Vq1atVKtWrXUrVs3pzjdNWHCBLVq1UqhoaFq3ry5Jk2apJKSknPGvfDCC4qNjVVoaKhuu+025efnO+1/+eWX1aZNG4WEhKh169b65z//6XEsAPyHZACGUatWLRUXFzvWN27cqKysLKWlpWnNmjUqKSlRcnKywsLCtGXLFn322WeqU6eOevbs6TjumWee0dKlS7V48WJ9+umnOn78uN59993f/dxhw4bpzTff1Pz587Vv3z698MILqlOnjmJjY/X2229LkrKyspSTk6PnnntOkjRz5ky99tprWrRokfbu3auxY8dq6NCh2rx5s6SzSUv//v3Vp08fZWZm6q677tLjjz/u8Z9JWFiYli5dqm+++UbPPfecXnrpJT377LNOY/bv369Vq1bpgw8+0Lp16/TVV1/p/vvvd+xftmyZJk+erKefflr79u3TjBkzNGnSJL366qsexwPAT+xAAEpJSbH37dvXbrfb7eXl5fa0tDR7cHCwffz48Y79UVFR9qKiIscxr7/+uj0+Pt5eXl7u2FZUVGSvVauW/aOPPrLb7XZ7dHS0ffbs2Y79JSUl9saNGzs+y26322+88Ub7Qw89ZLfb7fasrCy7JHtaWtp549y0aZNdkv3EiROObWfOnLGHhobat27d6jR25MiR9kGDBtntdrt94sSJ9rZt2zrtnzBhwjnn+i1J9nfffdfl/jlz5tg7d+7sWJ8yZYo9KCjIfuTIEce2f//733az2WzPycmx2+12e4sWLezLly93Os/06dPtiYmJdrvdbj948KBdkv2rr75y+bkA/Is5AwhYa9asUZ06dVRSUqLy8nINHjxYqampjv3t27d3miewa9cu7d+/X2FhYU7nOXPmjA4cOKD8/Hzl5OQoISHBsa9GjRq6+uqrz2kVVMjMzFRQUJBuvPFGt+Pev3+/Tp8+rZtuuslpe3Fxsa688kpJ0r59+5zikKTExES3P6PCypUrNX/+fB04cEAFBQUqLS1VeHi405gmTZrosssuc/qc8vJyZWVlKSwsTAcOHNDIkSM1atQox5jS0lJFRER4HA8A/yAZQMDq1q2bFi5cKIvFopiYGNWo4fzXvXbt2k7rBQUF6ty5s5YtW3bOuRo1alSpGGrVquXxMQUFBZKkDz/80OlLWDo7D8JXMjIyNGTIEE2dOlXJycmKiIjQihUr9Mwzz3gc60svvXROchIUFOSzWAFULZIBBKzatWurZcuWbo+/6qqrtHLlSkVGRp7z23GF6Ohobdu2TV26dJF09jfgHTt26Kqrrjrv+Pbt26u8vFybN29Wjx49ztlfUZkoKytzbGvbtq2Cg4OVnZ3tsqLQpk0bx2TICp9//vmFL/JXtm7dqri4OD3xxBOObT/88MM547Kzs3X06FHFxMQ4PsdsNis+Pl5RUVGKiYnR999/ryFDhnj0+QAuHkwgBP5nyJAhatiwofr27astW7bo4MGD+uSTT/Tggw/qyJEjkqSHHnpIs2bN0urVq/Xtt9/q/vvv/91nBDRt2lQpKSkaMWKEVq9e7TjnqlWrJElxcXEymUxas2aNfvrpJxUUFCgsLEzjx4/X2LFj9eqrr+rAgQPauXOnnn/+ecekvHvvvVffffedHn30UWVlZWn58uVaunSpR9d7+eWXKzs7WytWrNCBAwc0f/78806GDAkJUUpKinbt2qUtW7bowQcf1G233Sar1SpJmjp1qmbOnKn58+frP//5j/bs2aMlS5Zo7ty5HsUDwH9IBoD/CQ0NVXp6upo0aaL+/furTZs2GjlypM6cOeOoFDzyyCO64447lJKSosTERIWFhenPf/7z75534cKFuvXWW3X//ferdevWGjVqlAoLCyVJl112maZOnarHH39cUVFRGjNmjCRp+vTpmjRpkmbOnKk2bdqoZ8+e+vDDD9WsWTNJZ/v4b7/9tlavXq2OHTtq0aJFmjFjhkfXe8stt2js2LEaM2aMOnXqpK1bt2rSpEnnjGvZsqX69++vm2++WUlJSerQoYPTrYN33XWXXn75ZS1ZskTt27fXjTfeqKVLlzpiBXDxM9ldzXwCAACGQGUAAACDIxkAAMDgSAYAADA4kgEAAAyOZAAAAIMjGQAAwOBIBgAAMDiSAQAADI5kAAAAgyMZAADA4EgGAAAwuP8P2JqmLxnALnAAAAAASUVORK5CYII=", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "y_pred_train = rf.predict(X_train_scaled)\n", "\n", "print('F1 Score - Train Set : ', f1_score(y_train, y_pred_train), '\\n')\n", "print('Classification Report : \\n', classification_report(y_train, y_pred_train), '\\n')\n", "print('Confusion Matrix : \\n', ConfusionMatrixDisplay.from_estimator(rf, X_train_scaled, y_train, cmap='Reds'))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Dari classification report, kelihatan values accuracy, precision, dan recall sangat tinggi dan adalah overfitting. Kita bisa coba paramaeter tuning untuk memperbaik nilainya. " ] }, { "cell_type": "code", "execution_count": 407, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "F1 Score - All - Cross Validation : [0.49618321 0.41732283 0.4911032 ]\n", "F1 Score - Mean - Cross Validation : 0.46820308119983817\n", "F1 Score - Std - Cross Validation : 0.036037491823006584\n", "F1 Score - Range of Test-Set : 0.4321655893768316 - 0.5042405730228448\n" ] } ], "source": [ "\n", "f1_train_cross_val = cross_val_score(rf,\n", " X_train_scaled,\n", " y_train,\n", " cv=3,\n", " scoring=\"f1\")\n", "\n", "print('F1 Score - All - Cross Validation : ', f1_train_cross_val)\n", "print('F1 Score - Mean - Cross Validation : ', f1_train_cross_val.mean())\n", "print('F1 Score - Std - Cross Validation : ', f1_train_cross_val.std())\n", "print('F1 Score - Range of Test-Set : ', (f1_train_cross_val.mean()-f1_train_cross_val.std()) , '-', (f1_train_cross_val.mean()+f1_train_cross_val.std()))" ] }, { "cell_type": "code", "execution_count": 408, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "F1 Score - Test Set : 0.5471698113207547 \n", "\n", "Classification Report : \n", " precision recall f1-score support\n", "\n", " 0 0.87 0.94 0.90 467\n", " 1 0.67 0.46 0.55 126\n", "\n", " accuracy 0.84 593\n", " macro avg 0.77 0.70 0.72 593\n", "weighted avg 0.83 0.84 0.83 593\n", " \n", "\n", "Confusion Matrix : \n", " \n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "y_pred_test = rf.predict(X_test_scaled)\n", "\n", "print('F1 Score - Test Set : ', f1_score(y_test, y_pred_test), '\\n')\n", "print('Classification Report : \\n', classification_report(y_test, y_pred_test), '\\n')\n", "print('Confusion Matrix : \\n', ConfusionMatrixDisplay.from_estimator(rf, X_test_scaled, y_test, cmap='Reds'))" ] }, { "cell_type": "code", "execution_count": 409, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " Baseline (Default Hyperparameter)\n", "train - precision 0.771084\n", "train - recall 0.377210\n", "train - accuracy 0.842327\n", "train - f1_score 0.506596\n", "test - precision 0.769231\n", "test - recall 0.396825\n", "test - accuracy 0.846543\n", "test - f1_score 0.523560\n" ] } ], "source": [ "\n", "all_reports = {}\n", "\n", "def performance_report(all_reports, y_train, y_pred_train_svc, y_test, y_pred_test_svc, name):\n", " score_reports = {\n", " 'train - precision': precision_score(y_train, y_pred_train_svc),\n", " 'train - recall': recall_score(y_train, y_pred_train_svc),\n", " 'train - accuracy': accuracy_score(y_train, y_pred_train_svc),\n", " 'train - f1_score': f1_score(y_train, y_pred_train_svc),\n", " 'test - precision': precision_score(y_test, y_pred_test_svc),\n", " 'test - recall': recall_score(y_test, y_pred_test_svc),\n", " 'test - accuracy': accuracy_score(y_test, y_pred_test_svc),\n", " 'test - f1_score': f1_score(y_test, y_pred_test_svc),\n", " }\n", " all_reports[name] = score_reports\n", " return all_reports\n", "\n", "# Example usage (ensure y_pred_train_svc and y_pred_test_svc are defined)\n", "all_reports = performance_report(all_reports, y_train, y_pred_train_svc, y_test, y_pred_test_svc, 'Baseline (Default Hyperparameter)')\n", "report_df = pd.DataFrame(all_reports)\n", "print(report_df)\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan coba save baseline model kedalam dictionary \"all_report\" untuk menumpang data dari random dan grid search. " ] }, { "cell_type": "code", "execution_count": 410, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/opt/miniconda3/lib/python3.12/site-packages/threadpoolctl.py:1214: RuntimeWarning: \n", "Found Intel OpenMP ('libiomp') and LLVM OpenMP ('libomp') loaded at\n", "the same time. Both libraries are known to be incompatible and this\n", "can cause random crashes or deadlocks on Linux when loaded in the\n", "same Python program.\n", "Using threadpoolctl may cause crashes or deadlocks. For more\n", "information and possible workarounds, please see\n", " https://github.com/joblib/threadpoolctl/blob/master/multiple_openmp.md\n", "\n", " warnings.warn(msg, RuntimeWarning)\n", "/opt/miniconda3/lib/python3.12/site-packages/threadpoolctl.py:1214: RuntimeWarning: \n", "Found Intel OpenMP ('libiomp') and LLVM OpenMP ('libomp') loaded at\n", "the same time. Both libraries are known to be incompatible and this\n", "can cause random crashes or deadlocks on Linux when loaded in the\n", "same Python program.\n", "Using threadpoolctl may cause crashes or deadlocks. For more\n", "information and possible workarounds, please see\n", " https://github.com/joblib/threadpoolctl/blob/master/multiple_openmp.md\n", "\n", " warnings.warn(msg, RuntimeWarning)\n", "/opt/miniconda3/lib/python3.12/site-packages/threadpoolctl.py:1214: RuntimeWarning: \n", "Found Intel OpenMP ('libiomp') and LLVM OpenMP ('libomp') loaded at\n", "the same time. Both libraries are known to be incompatible and this\n", "can cause random crashes or deadlocks on Linux when loaded in the\n", "same Python program.\n", "Using threadpoolctl may cause crashes or deadlocks. For more\n", "information and possible workarounds, please see\n", " https://github.com/joblib/threadpoolctl/blob/master/multiple_openmp.md\n", "\n", " warnings.warn(msg, RuntimeWarning)\n", "/opt/miniconda3/lib/python3.12/site-packages/threadpoolctl.py:1214: RuntimeWarning: \n", "Found Intel OpenMP ('libiomp') and LLVM OpenMP ('libomp') loaded at\n", "the same time. Both libraries are known to be incompatible and this\n", "can cause random crashes or deadlocks on Linux when loaded in the\n", "same Python program.\n", "Using threadpoolctl may cause crashes or deadlocks. For more\n", "information and possible workarounds, please see\n", " https://github.com/joblib/threadpoolctl/blob/master/multiple_openmp.md\n", "\n", " warnings.warn(msg, RuntimeWarning)\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Best Parameters: {'bootstrap': False, 'max_depth': 30, 'max_features': 'sqrt', 'min_samples_leaf': 7, 'min_samples_split': 8, 'n_estimators': 159}\n", "Best Score: 0.4781662991940268\n", " precision recall f1-score support\n", "\n", " 0 0.86 0.96 0.91 467\n", " 1 0.73 0.44 0.55 126\n", "\n", " accuracy 0.85 593\n", " macro avg 0.80 0.70 0.73 593\n", "weighted avg 0.84 0.85 0.83 593\n", "\n" ] } ], "source": [ "\n", "# Define the parameter distribution\n", "random_search_params = {\n", " 'n_estimators': randint(50, 200),\n", " 'max_features': ['auto', 'sqrt', 'log2'],\n", " 'max_depth': [None] + list(range(10, 110, 10)),\n", " 'min_samples_split': randint(2, 11),\n", " 'min_samples_leaf': randint(1, 11),\n", " 'bootstrap': [True, False]\n", "}\n", "\n", "\n", "# Initialize RandomizedSearchCV with reduced iterations and fewer CV folds\n", "rf_randomcv = RandomizedSearchCV(estimator=RandomForestClassifier(),\n", " param_distributions=random_search_params,\n", " n_iter=20, # Reduced number of iterations\n", " cv=3, # Reduced number of CV folds\n", " random_state=46,\n", " n_jobs=-1,\n", " scoring='f1')\n", "\n", "# Fit the model\n", "rf_randomcv.fit(X_train_scaled, y_train)\n", "\n", "# Print the best parameters and the best score\n", "print(\"Best Parameters:\", rf_randomcv.best_params_)\n", "print(\"Best Score:\", rf_randomcv.best_score_)\n", "\n", "# Make predictions with the best estimator\n", "y_pred = rf_randomcv.predict(X_test_scaled)\n", "print(classification_report(y_test, y_pred))\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan membuat parameter untuk random search dengan model RandomizedSearchCV untuk mengoptimisasi performance model. Selection dari n_iter atau CV bisa diganti untuk ganti random combination yang akan digunakan dalam search process, dan juga mengoptimisasi beberapa K-folds yang akan di validate dalam beberapa set. \n", "\n" ] }, { "cell_type": "code", "execution_count": 411, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "{'bootstrap': False,\n", " 'max_depth': 30,\n", " 'max_features': 'sqrt',\n", " 'min_samples_leaf': 7,\n", " 'min_samples_split': 8,\n", " 'n_estimators': 159}" ] }, "execution_count": 411, "metadata": {}, "output_type": "execute_result" } ], "source": [ "rf_randomcv.best_params_" ] }, { "cell_type": "code", "execution_count": 412, "metadata": {}, "outputs": [], "source": [ "rf_randomcv_best = rf_randomcv.best_estimator_" ] }, { "cell_type": "code", "execution_count": 413, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "F1 Score - Test Set : 0.5517241379310345 \n", "\n", "Classification Report : \n", " precision recall f1-score support\n", "\n", " 0 0.86 0.96 0.91 467\n", " 1 0.73 0.44 0.55 126\n", "\n", " accuracy 0.85 593\n", " macro avg 0.80 0.70 0.73 593\n", "weighted avg 0.84 0.85 0.83 593\n", " \n", "\n", "Confusion Matrix : \n", " \n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Check Performance Model against Test-Set\n", "\n", "y_pred_test = rf_randomcv_best.predict(X_test_scaled)\n", "\n", "print('F1 Score - Test Set : ', f1_score(y_test, y_pred_test), '\\n')\n", "print('Classification Report : \\n', classification_report(y_test, y_pred_test), '\\n')\n", "print('Confusion Matrix : \\n', ConfusionMatrixDisplay.from_estimator(rf_randomcv_best, X_test_scaled, y_test, cmap='Reds'))" ] }, { "cell_type": "code", "execution_count": 414, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Baseline (Default Hyperparameter)Random Search
train - precision0.7710840.771084
train - recall0.3772100.377210
train - accuracy0.8423270.842327
train - f1_score0.5065960.506596
test - precision0.7692310.769231
test - recall0.3968250.396825
test - accuracy0.8465430.846543
test - f1_score0.5235600.523560
\n", "
" ], "text/plain": [ " Baseline (Default Hyperparameter) Random Search\n", "train - precision 0.771084 0.771084\n", "train - recall 0.377210 0.377210\n", "train - accuracy 0.842327 0.842327\n", "train - f1_score 0.506596 0.506596\n", "test - precision 0.769231 0.769231\n", "test - recall 0.396825 0.396825\n", "test - accuracy 0.846543 0.846543\n", "test - f1_score 0.523560 0.523560" ] }, "execution_count": 414, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Save Classification Report into a Dictionary\n", "\n", "all_reports = performance_report(all_reports, y_train, y_pred_train_svc, y_test, y_pred_test_svc, 'Random Search')\n", "pd.DataFrame(all_reports)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Bisa kelihatan dari report bahwa tidak ada perbedaan dari baseline dan random search. Kita akan coba save model kedalam dictionary. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Grid Search" ] }, { "cell_type": "code", "execution_count": 415, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "{'n_estimators': [100, 200],\n", " 'max_features': ['auto', 'sqrt'],\n", " 'max_depth': [10, 20, None],\n", " 'min_samples_split': [2, 5],\n", " 'min_samples_leaf': [1, 2],\n", " 'bootstrap': [True, False]}" ] }, "execution_count": 415, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Define a simplified parameter grid\n", "grid_search_params = {\n", " 'n_estimators': [100, 200],\n", " 'max_features': ['auto', 'sqrt'],\n", " 'max_depth': [10, 20, None],\n", " 'min_samples_split': [2, 5],\n", " 'min_samples_leaf': [1, 2],\n", " 'bootstrap': [True, False]\n", "}\n", "\n", "grid_search_params\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan set parameter untuk grid search dengan variables n_estimators, max_features, dan lain lain. " ] }, { "cell_type": "code", "execution_count": 416, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "{'bootstrap': False,\n", " 'max_depth': 30,\n", " 'max_features': 'sqrt',\n", " 'min_samples_leaf': 7,\n", " 'min_samples_split': 8,\n", " 'n_estimators': 159}" ] }, "execution_count": 416, "metadata": {}, "output_type": "execute_result" } ], "source": [ "rf_randomcv.best_params_" ] }, { "cell_type": "code", "execution_count": 417, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Fitting 3 folds for each of 96 candidates, totalling 288 fits\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.9s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.9s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 1.9s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.9s\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.4s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 2.4s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.5s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 2.5s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 2.4s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 2.3s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 2.3s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.2s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 2.7s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 2.7s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.8s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 3.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.7s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.4s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 2.6s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.8s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 3.1s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 3.4s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 3.3s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.1s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=True, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 2.2s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.4s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 2.7s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.2s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 2.6s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 2.6s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.1s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 2.4s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 2.3s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 2.3s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 2.3s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.9s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.0s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 1.9s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 1.9s\n", "[CV] END bootstrap=True, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 1.9s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 2.5s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 2.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 1.6s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 1.4s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.6s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 1.4s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.7s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.6s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 1.2s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.7s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 1.4s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 3.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 2.1s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 3.2s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 3.5s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.9s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.7s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 3.2s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.5s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 3.3s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.4s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 2.7s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.2s\n", "[CV] END bootstrap=True, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 2.7s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.2s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.5s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 2.9s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 2.8s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 2.9s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 2.5s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 2.4s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 2.1s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.7s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 1.3s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 1.1s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 1.0s\n", "[CV] END bootstrap=False, max_depth=10, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 1.0s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.7s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.7s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.6s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.6s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 1.3s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.6s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.6s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 1.3s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.6s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 1.3s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 1.4s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 1.7s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.1s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.2s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.2s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 2.4s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 2.3s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.2s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 2.4s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=auto, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 0.0s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 2.2s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 3.4s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 3.3s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 3.4s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=100; total time= 1.2s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.3s\n", "[CV] END bootstrap=False, max_depth=20, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 2.3s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 2.7s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.5s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=100; total time= 1.8s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 3.2s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.4s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 3.5s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 3.3s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.6s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 3.2s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=100; total time= 1.6s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 3.4s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 3.3s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 3.6s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=2, n_estimators=200; total time= 3.5s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.5s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.6s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=100; total time= 1.4s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 2.3s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 2.2s\n", "[CV] END bootstrap=False, max_depth=None, max_features=sqrt, min_samples_leaf=2, min_samples_split=5, n_estimators=200; total time= 2.1s\n" ] }, { "data": { "text/html": [ "
GridSearchCV(cv=3, estimator=RandomForestClassifier(), n_jobs=-1,\n",
       "             param_grid={'bootstrap': [True, False],\n",
       "                         'max_depth': [10, 20, None],\n",
       "                         'max_features': ['auto', 'sqrt'],\n",
       "                         'min_samples_leaf': [1, 2],\n",
       "                         'min_samples_split': [2, 5],\n",
       "                         'n_estimators': [100, 200]},\n",
       "             scoring='f1', verbose=2)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "GridSearchCV(cv=3, estimator=RandomForestClassifier(), n_jobs=-1,\n", " param_grid={'bootstrap': [True, False],\n", " 'max_depth': [10, 20, None],\n", " 'max_features': ['auto', 'sqrt'],\n", " 'min_samples_leaf': [1, 2],\n", " 'min_samples_split': [2, 5],\n", " 'n_estimators': [100, 200]},\n", " scoring='f1', verbose=2)" ] }, "execution_count": 417, "metadata": {}, "output_type": "execute_result" } ], "source": [ "rf_gridcv = GridSearchCV(estimator=RandomForestClassifier(),\n", " param_grid=grid_search_params,\n", " cv=3,\n", " n_jobs=-1,\n", " verbose=2,\n", " scoring='f1')\n", "\n", "rf_gridcv.fit(X_train_scaled, y_train)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan coba grid search dengan cross validation yang ditepatkan. " ] }, { "cell_type": "code", "execution_count": 418, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "{'bootstrap': True,\n", " 'max_depth': None,\n", " 'max_features': 'sqrt',\n", " 'min_samples_leaf': 2,\n", " 'min_samples_split': 2,\n", " 'n_estimators': 100}" ] }, "execution_count": 418, "metadata": {}, "output_type": "execute_result" } ], "source": [ "rf_gridcv.best_params_" ] }, { "cell_type": "code", "execution_count": 419, "metadata": {}, "outputs": [], "source": [ "rf_gridcv_best = rf_gridcv.best_estimator_" ] }, { "cell_type": "code", "execution_count": 420, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "F1 Score - Test Set : 0.5507246376811594 \n", "\n", "Classification Report : \n", " precision recall f1-score support\n", "\n", " 0 0.87 0.95 0.91 467\n", " 1 0.70 0.45 0.55 126\n", "\n", " accuracy 0.84 593\n", " macro avg 0.78 0.70 0.73 593\n", "weighted avg 0.83 0.84 0.83 593\n", " \n", "\n", "Confusion Matrix : \n", " \n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "y_pred_test = rf_gridcv_best.predict(X_test_scaled)\n", "\n", "print('F1 Score - Test Set : ', f1_score(y_test, y_pred_test), '\\n')\n", "print('Classification Report : \\n', classification_report(y_test, y_pred_test), '\\n')\n", "print('Confusion Matrix : \\n', ConfusionMatrixDisplay.from_estimator(rf_gridcv_best, X_test_scaled, y_test, cmap='Reds'))" ] }, { "cell_type": "code", "execution_count": 421, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Baseline (Default Hyperparameter)Random SearchGrid Search
train - precision0.7710840.7710840.771084
train - recall0.3772100.3772100.377210
train - accuracy0.8423270.8423270.842327
train - f1_score0.5065960.5065960.506596
test - precision0.7692310.7692310.769231
test - recall0.3968250.3968250.396825
test - accuracy0.8465430.8465430.846543
test - f1_score0.5235600.5235600.523560
\n", "
" ], "text/plain": [ " Baseline (Default Hyperparameter) Random Search \\\n", "train - precision 0.771084 0.771084 \n", "train - recall 0.377210 0.377210 \n", "train - accuracy 0.842327 0.842327 \n", "train - f1_score 0.506596 0.506596 \n", "test - precision 0.769231 0.769231 \n", "test - recall 0.396825 0.396825 \n", "test - accuracy 0.846543 0.846543 \n", "test - f1_score 0.523560 0.523560 \n", "\n", " Grid Search \n", "train - precision 0.771084 \n", "train - recall 0.377210 \n", "train - accuracy 0.842327 \n", "train - f1_score 0.506596 \n", "test - precision 0.769231 \n", "test - recall 0.396825 \n", "test - accuracy 0.846543 \n", "test - f1_score 0.523560 " ] }, "execution_count": 421, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Save Classification Report into a Dictionary\n", "\n", "all_reports = performance_report(all_reports, y_train, y_pred_train_svc, y_test, y_pred_test_svc, 'Grid Search')\n", "pd.DataFrame(all_reports)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Bisa kelihatan dari report bahwa tidak ada perbedaan dari baseline, random search dan grid search. Kita akan coba save model kedalam dictionary. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Model Saving" ] }, { "cell_type": "code", "execution_count": 422, "metadata": {}, "outputs": [], "source": [ "#Model saving\n", "\n", "with open('list_num_cols.txt', 'w') as file_1:\n", " json.dump(selected_num_cols, file_1)\n", "\n", "with open('list_cat_cols.txt', 'w') as file_2:\n", " json.dump(selected_cat_cols, file_2)\n", "\n", "with open('scaler.pkl', 'wb') as file_3:\n", " pickle.dump(scaler, file_3)\n", "\n", "#with open('encoder.pkl', 'wb') as file_4:\n", " #pickle.dump(encoder, file_4)\n", "\n", "with open('model.pkl', 'wb') as file_5:\n", " pickle.dump(lg, file_5)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan save data-data tersebut didalam Json files." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Model Inference " ] }, { "cell_type": "code", "execution_count": 423, "metadata": {}, "outputs": [], "source": [ "# Load model and other files\n", "\n", "with open('list_cat_cols.txt', 'r') as file_1:\n", " list_cat_col = json.load(file_1)\n", "\n", "with open('list_num_cols.txt', 'r') as file_2:\n", " list_num_col = json.load(file_2)\n", "\n", "with open(\"model.pkl\", \"rb\") as file_3:\n", " model = pickle.load(file_3)\n", "\n", "with open(\"scaler.pkl\", \"rb\") as file_4:\n", " scaler = pickle.load(file_4)\n", "\n", "#with open(\"encoder.pkl\", \"rb\") as file_5:\n", " #encoder = pickle.load(file_5)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan coba buka yang kita tadi save untuk dipake untuk inference." ] }, { "cell_type": "code", "execution_count": 424, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " limit_balance sex education_level marital_status age pay_0 pay_2 \\\n", "0 80000.0 1 6 1 54.0 0.0 0.0 \n", "1 200000.0 1 4 1 49.0 0.0 0.0 \n", "2 20000.0 2 6 2 22.0 0.0 0.0 \n", "3 260000.0 2 4 2 33.0 0.0 0.0 \n", "4 150000.0 1 4 2 32.0 0.0 0.0 \n", "\n", " pay_3 pay_4 pay_5 ... bill_amt_4 bill_amt_5 bill_amt_6 pay_amt_1 \\\n", "0 0.0 0.0 0.0 ... 29296.0 26210.0 17643.0 2545.0 \n", "1 0.0 0.0 0.0 ... 50146.0 50235.0 48984.0 1689.0 \n", "2 0.0 0.0 0.0 ... 1434.0 500.0 0.0 4641.0 \n", "3 0.0 0.0 0.0 ... 27821.0 30767.0 29890.0 5000.0 \n", "4 0.0 -1.0 0.0 ... 150464.0 143375.0 146411.0 4019.0 \n", "\n", " pay_amt_2 pay_amt_3 pay_amt_4 pay_amt_5 pay_amt_6 \\\n", "0 2208.0 1336.0 2232.0 542.0 348.0 \n", "1 2164.0 2500.0 3480.0 2500.0 3000.0 \n", "2 1019.0 900.0 0.0 1500.0 0.0 \n", "3 5000.0 1137.0 5000.0 1085.0 5000.0 \n", "4 146896.0 157436.0 4600.0 4709.0 5600.0 \n", "\n", " default_payment_next_month \n", "0 1 \n", "1 0 \n", "2 1 \n", "3 0 \n", "4 0 \n", "\n", "[5 rows x 24 columns]\n" ] } ], "source": [ "import pandas as pd\n", "\n", "# Read the data from the CSV file\n", "file_path = \"/Users/ryantrisnadi/Desktop/first_project1/p1-ftds017-hck-g5-ryantrisnadi/_P1G5_Set_1_Ryan_Trisnadi.csv\"\n", "df_original = pd.read_csv(file_path)\n", "\n", "# Create a new DataFrame with the specified index columns\n", "index_columns = ['limit_balance', 'sex', 'education_level', 'marital_status', 'age',\n", " 'pay_0', 'pay_2', 'pay_3', 'pay_4', 'pay_5', 'pay_6', 'bill_amt_1',\n", " 'bill_amt_2', 'bill_amt_3', 'bill_amt_4', 'bill_amt_5', 'bill_amt_6',\n", " 'pay_amt_1', 'pay_amt_2', 'pay_amt_3', 'pay_amt_4', 'pay_amt_5',\n", " 'pay_amt_6', 'default_payment_next_month']\n", "df_data_dummy = df_original[index_columns].copy()\n", "\n", "# Display the new DataFrame\n", "print(df_data_dummy.head())\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita akan menggunakan data asli yang belum diolah sebagai data dummy dinamakan \"data1\"" ] }, { "cell_type": "code", "execution_count": 425, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "RangeIndex: 2965 entries, 0 to 2964\n", "Data columns (total 24 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 limit_balance 2965 non-null float64\n", " 1 sex 2965 non-null int64 \n", " 2 education_level 2965 non-null int64 \n", " 3 marital_status 2965 non-null int64 \n", " 4 age 2965 non-null float64\n", " 5 pay_0 2965 non-null float64\n", " 6 pay_2 2965 non-null float64\n", " 7 pay_3 2965 non-null float64\n", " 8 pay_4 2965 non-null float64\n", " 9 pay_5 2965 non-null float64\n", " 10 pay_6 2965 non-null float64\n", " 11 bill_amt_1 2965 non-null float64\n", " 12 bill_amt_2 2965 non-null float64\n", " 13 bill_amt_3 2965 non-null float64\n", " 14 bill_amt_4 2965 non-null float64\n", " 15 bill_amt_5 2965 non-null float64\n", " 16 bill_amt_6 2965 non-null float64\n", " 17 pay_amt_1 2965 non-null float64\n", " 18 pay_amt_2 2965 non-null float64\n", " 19 pay_amt_3 2965 non-null float64\n", " 20 pay_amt_4 2965 non-null float64\n", " 21 pay_amt_5 2965 non-null float64\n", " 22 pay_amt_6 2965 non-null float64\n", " 23 default_payment_next_month 2965 non-null int64 \n", "dtypes: float64(20), int64(4)\n", "memory usage: 556.1 KB\n" ] } ], "source": [ "df_data_dummy.info()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Cek data \"dummy\" jika sudah benar dengan yang original. " ] }, { "cell_type": "code", "execution_count": 426, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Index(['limit_balance', 'sex', 'education_level', 'marital_status', 'age',\n", " 'pay_0', 'pay_2', 'pay_3', 'pay_4', 'pay_5', 'pay_6', 'bill_amt_1',\n", " 'bill_amt_2', 'bill_amt_3', 'bill_amt_4', 'bill_amt_5', 'bill_amt_6',\n", " 'pay_amt_1', 'pay_amt_2', 'pay_amt_3', 'pay_amt_4', 'pay_amt_5',\n", " 'pay_amt_6', 'default_payment_next_month'],\n", " dtype='object')" ] }, "execution_count": 426, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_data_dummy.columns" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Cek kolom dari yang \"dummy\" jika sudah sesuai." ] }, { "cell_type": "code", "execution_count": 427, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[[8.86075949e-02 2.00000000e-01 2.22222222e-01 ... 1.39665164e-03\n", " 8.62453532e-04 6.00000000e+00]\n", " [2.40506329e-01 2.00000000e-01 2.22222222e-01 ... 6.44212013e-03\n", " 7.43494424e-03 4.00000000e+00]\n", " [1.26582278e-02 2.00000000e-01 2.22222222e-01 ... 3.86527208e-03\n", " 0.00000000e+00 6.00000000e+00]\n", " ...\n", " [5.56962025e-01 0.00000000e+00 0.00000000e+00 ... 1.00497074e-03\n", " 9.66542751e-04 2.00000000e+00]\n", " [5.06329114e-02 0.00000000e+00 0.00000000e+00 ... 0.00000000e+00\n", " 1.93308550e-03 2.00000000e+00]\n", " [3.54430380e-01 3.00000000e-01 0.00000000e+00 ... 1.00497074e-03\n", " 1.63990087e-02 2.00000000e+00]]\n" ] } ], "source": [ "\n", "# Assuming df_data_dummy, selected_num_cols, and selected_cat_cols are defined\n", "data_inference_num = df_data_dummy[selected_num_cols]\n", "data_inference_cat = df_data_dummy[selected_cat_cols]\n", "\n", "data_inference_num.fillna(0, inplace=True) # Assuming missing values are filled with 0\n", "data_inference_cat.fillna('Unknown', inplace=True) # Filling categorical missing values with 'Unknown'\n", "\n", "# Fit the MinMaxScaler on the numerical features\n", "scaler.fit(data_inference_num)\n", "\n", "# Transform the numerical features using the fitted scaler\n", "data_inference_num_scaled = scaler.transform(data_inference_num)\n", "\n", "# Concatenate the scaled numerical features and the categorical features\n", "data_inference_final = np.concatenate([data_inference_num_scaled, data_inference_cat], axis=1)\n", "\n", "# Now data_inference_final contains both scaled numerical features and categorical features\n", "print(data_inference_final)\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Kita mau bagi dari data dummy dengan list numerical dan categorical dari hasil \"data\" original. Kita akan bagi jadi dua variable dan masukan mengunakan scalar jadi bisa membuat inference baru jika digabung lagi dengan feature \"concatenate\". Kita bisa terus melihat prediksi output dengan data \"dummy\" dengan target variable. " ] }, { "cell_type": "code", "execution_count": 428, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "['education_level']\n", "['limit_balance', 'pay_0', 'pay_2', 'pay_3', 'pay_4', 'pay_5', 'pay_6', 'pay_amt_1', 'pay_amt_2', 'pay_amt_3', 'pay_amt_4', 'pay_amt_5', 'pay_amt_6']\n" ] } ], "source": [ "print(selected_cat_cols)\n", "print(selected_num_cols)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Lihat kolom yang akan digunakan karena terpilih sebagai signifikan data." ] }, { "cell_type": "code", "execution_count": 429, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Predicted Overall: 0.00\n" ] } ], "source": [ "# Predict the score\n", "predicted_score = lg.predict(data_inference_final)\n", "\n", "# Show result\n", "print(f\"Predicted Overall: {predicted_score[0]:.2f}\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Insight: Dari data inferensial, Kita bisa lihat bahwa jika faktor variable sudah semua dimasukan kedalam model, maka output dari target variable akan keluarkan 0.00. Ini artinya data dummy dan original tidak sesuai untuk model tersebut. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Conclusion" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Setelah kita menggunakan test Logistic Regression, KNN, dan SVM, kita bisa lihat perbandingan model yang digunakan oleh machine learning untuk prediksi bagaimana client bisa diprediksi untuk default payment atau tidak. Karena saking banyak numerik data yang harus diolah, maka interpretasi dari setiap test harus bisa disimpulkan dalam parameter dan cross validation. Di konteks test ini, SVM telah dipakai karena skor dari akurasi, recall, precision, dan F-1 rata-rata yang paling tinggi, tetapi ini tidak selalu benar dalam setiap situasi. Bisa bilang saja ada dataset yang memiliki variable age, sex, marital_status, atau education_level yang beda saja akan mempunyai dampak besar terhadap test regresi atau model yang digunakan untuk KNN atau SVM.\n", "\n", "Kita menggunakan SVM di cross validation dan parameter tuning di test ini karena leih effektif model dalam dimensi lebih tinggi karena klassifikasi fitur sangat tinggi. Juga, kita bisa pilih antar linear, polynomial, dan RBF untuk pilih perbedaan kompleks dataset dengan boundaries yang linear atau non-linear. Oleh karena itu, Data yang digunakan bisa di \"customize\" terhadap splitting dataset yang digunakan karena sebagai analis, kita tidak tau distribusi dari setiap sample size, dan SVM adalah model yang paling flexible untuk menjelaskan relasi antar \"default_payment\" terhadap variable numerikal dan kategorikal di dataset tersebut. \n", "\n", "Dalam kesimpulan, bisa dikatakan bahwa faktor limit_balance, education_level, pay, dan pay_amt memiliki dampak paling signifikan terhadap prediksi jika client akan lunas bayar tagihan. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Pertanyaan:\n", "\n", " Resiko apa saja client bank dengan limit balance yang sangat tinggi terhadap utang mereka?\n", " Client yang memiliki limit balance lebih tinggi cenderung bayar tagihan pada tepat waktu daripada yang gagal bayar. \n", "\n", " Apakah ada dampak \"education\" terhadap default_payment_next_month?\n", " Pendidikan client yang lebih tinggi memiliki ratio yang lebih baik terhadap kelunasan pembayaran utang. \n", "\n", " Adakah perbedaan \"pay\" dan \"bill_amt\" yang bisa disimpulkan terhadap kemampuan client untuk membayar utang balik? \n", " Client yang memiliki pay paling dekat terhadap bill_amt akan punya kemungkinan lebih besar untuk bayar balik utang pada bulan itu, dan tidak akan \"default\" untuk bulan keberikutnya. \n", " \n", " Resiko yang paling tinggi di kolom-kolom apa saja?\n", " Limit balance, age, dan education level dan adalah faktor yang memiliki dampak terbesar terhadap \"default\" payment dari client. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Conceptual Problems\n", "### Jawab pertanyaan berikut:\n", "\n", " Apakah yang dimaksud dengan coeficient pada logistic regression?\n", " Coefficient pada Logistic Regression mengacu pada nilai yang menunjukkan seberapa besar pengaruh suatu fitur terhadap hasil prediksi model. Dalam konteks logistic regression, koefisien digunakan untuk mengkalibrasi kontribusi setiap fitur terhadap probabilitas prediksi kelas target yang diberikan. Koefisien positif menunjukkan korelasi positif antara fitur tersebut dengan kelas target, sedangkan koefisien negatif menunjukkan korelasi negatif.\n", "\n", " Apakah fungsi parameter kernel pada SVM? Jelaskan salah satu kernel yang kalian pahami!\n", " Parameter kernel pada Support Vector Machine (SVM) digunakan untuk menentukan jenis fungsi kernel yang digunakan untuk mentransformasikan data ke dalam ruang fitur yang lebih tinggi. Salah satu kernel yang umum digunakan adalah kernel RBF (Radial Basis Function). Kernel RBF memungkinkan SVM untuk menangani data yang tidak linear dengan memetakan data ke dalam ruang dimensi yang lebih tinggi, di mana pemisah linear menjadi lebih mungkin.\n", "\n", " Bagaimana cara memilih K yang optimal pada KNN?\n", " Untuk memilih nilai K yang optimal pada K-Nearest Neighbors (KNN), dapat dilakukan dengan menggunakan metode validasi silang (cross-validation). Dengan menggunakan teknik ini, kita dapat membagi data menjadi subset untuk pelatihan dan pengujian, dan kemudian menghitung performa model KNN dengan berbagai nilai K. Nilai K yang memberikan performa terbaik pada data pengujian adalah K yang optimal untuk digunakan pada model KNN.\n", "\n", " Apa yang dimaksud dengan metrics-metrics berikut : Accuracy, Precision, Recall, F1 Score, dan kapan waktu yang tepat untuk menggunakannya?\n", " Accuracy: Mengukur seberapa sering model melakukan prediksi yang benar dari semua prediksi yang dilakukan. Baik digunakan ketika kelas target memiliki distribusi yang seimbang.\n", " Precision: Mengukur proporsi prediksi positif yang benar dari total prediksi positif. Berguna ketika penting untuk meminimalkan false positive.\n", " Recall: Mengukur proporsi positif sebenarnya yang diprediksi dengan benar dari semua kelas positif yang sebenarnya. Berguna ketika penting untuk meminimalkan false negative.\n", " F1 Score: Kombinasi dari Precision dan Recall, digunakan untuk mengukur keseimbangan antara Precision dan Recall. Berguna ketika kelas target tidak seimbang.\n", " Waktu yang tepat untuk menggunakan masing-masing metrik ini tergantung pada tujuan dan karakteristik dari data serta masalah yang sedang dihadapi. Sebagai contoh, jika kelas target seimbang, maka Accuracy dapat menjadi metrik yang baik. Namun, jika kelas target tidak seimbang (imbalance class), maka menggunakan Precision, Recall, dan F1 Score dapat memberikan gambaran yang lebih akurat tentang performa model.\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Recommendation" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Di test tersebut ada beberapa masalah dengan merangka model seperti waktu di hypertuning test, CV atau n_iter terlalu kecil karena waktu untuk proses terlalu lama. Oleh karena itu test telah diganti variable CV atau n_iter yang lebih kecil untuk perpendek waktu test. Dibawah ada beberapa saran untuk memperbaik test diatas untuk lain kali:\n", "\n", "1. Waktu Cross validation dan hypertuning test, coba CV, n-iter, atau n_jobs yang lebih besar jadi bisa dapat akurasi, precision, recall, dan f-1 score yang lebih jelas dan beda dari yang original.\n", "\n", "2. Modifikasi data dummy dengan variable yang teratur jadi model inferensi memiliki jawaban yang bukan 0.00." ] } ], "metadata": { "kernelspec": { "display_name": "base", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.12.2" } }, "nbformat": 4, "nbformat_minor": 2 }