Upload 6 files

app.py

@@ -0,0 +1,14 @@
+# Import Libraries
+import streamlit as st
+
+# Import finished streamlit pages
+import eda
+import model
+
+# Navigation Button
+navi = st.sidebar.selectbox('Choose page: ', ('Predictor', 'EDA'))
+
+if navi == 'Predictor':
+    model.run()
+else:
+    eda.run()

eda.py

@@ -0,0 +1,61 @@
+# Import Essential Library
+import streamlit as st
+import pandas as pd
+
+# Library for Visualization
+import matplotlib.pyplot as plt
+import seaborn as sns
+
+# Function to run EDA
+def run():
+    # Set Title
+    st.title('Insurance Lead Prediction Model')
+
+    # Sub Title
+    st.subheader('Exploratory Data Analysis Section')
+    st.markdown('---')
+
+    # Insert Image
+    st.image('https://www.startinsland.de/site/assets/files/4129/tk-logo_koop_official_health_partner_pos.800x0.png')
+
+    # Markdown
+    st.markdown('# Dataframe Insurance Lead')
+
+    # Load Data
+    data = pd.read_csv('data_eda.csv')
+
+    # Display dataframe in StreamLit
+    st.dataframe(data.head(20))
+    st.markdown('---')
+
+    # EDA
+    st.markdown('## EDA')
+
+    # Convert Rate Balance Visualization
+    st.markdown('### Convert Rate Balance')
+    canvas = plt.figure(figsize=(10,5))
+    sns.barplot(x=data['Response'].value_counts().index, y=data['Response'].value_counts(), hue=data['Response'].value_counts().index)
+    st.pyplot(canvas)
+    st.markdown('Data is still slightly imbalanced (biased towards clients who will not likely convert)')
+
+    # Holding Policy Duration Distribution Visualization 
+    st.markdown('### Holding Policy Duration Distribution')
+    canvas = plt.figure(figsize=(10,5))
+    sns.histplot(data['Holding_Policy_Duration'], kde=True, bins=15)
+    st.pyplot(canvas)
+
+    # Holding Policy Type Distribution Visualization 
+    st.markdown('### Holding Policy Type Distribution')
+    canvas = plt.figure(figsize=(10,5))
+    sns.barplot(x=data['Holding_Policy_Type'].value_counts().index, y=data['Holding_Policy_Type'].value_counts(), hue=data['Holding_Policy_Type'].value_counts().index)
+    st.pyplot(canvas)
+
+    # Recommended Policy Category Distribution Visualization 
+    st.markdown('### Recommended Policy Category Distribution')
+    canvas = plt.figure(figsize=(15,5))
+    sns.barplot(x=data['Reco_Policy_Cat'].value_counts().index, y=data['Reco_Policy_Cat'].value_counts(), hue=data['Reco_Policy_Cat'].value_counts().index)
+    st.pyplot(canvas)
+
+
+if __name__=='__main__':
+    run()

model.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5901c17ef92b073b79655e7bc96cab27876d787492a28271179e25f46d16c02c
+size 34246

model.py

@@ -0,0 +1,54 @@
+# Import Essential Library
+import streamlit as st
+import pandas as pd
+import pickle
+
+# Load Model
+with open('model.pkl', 'rb') as file:
+    model = pickle.load(file)
+
+list_cat_cols = ['education_level', 'pay_sep05', 'pay_aug05', 'pay_jul05', 'pay_jun05', 'pay_may05', 'pay_apr05']
+list_num_cols = ['limit_balance', 'pay_amt_sep05', 'pay_amt_aug05', 'pay_amt_jul05', 'pay_amt_jun05', 'pay_amt_may05', 'pay_amt_apr05']
+
+# Function to run model predictor
+def run():
+    # Set Title
+    st.title('Credit Card Default Prediction Model')
+
+    # Sub Title
+    st.subheader('Model Predict Section')
+    st.markdown('---')
+
+    # Insert Image
+    st.image('https://www.startinsland.de/site/assets/files/4129/tk-logo_koop_official_health_partner_pos.800x0.png')
+
+    # Creating Form for Data Inference
+    st.markdown('## Input Data')
+    with st.form('my_form'):
+        Holding_Policy_Duration = st.slider('Holding Policy Duration', min_value=1, max_value=14, value=2, step=1)
+        Holding_Policy_Type = st.selectbox('Holding Policy Type', (1, 2, 3, 4))
+        Reco_Policy_Cat = st.slider('Recommended Policy Category', min_value=1, max_value=22, value=6, step=1)
+
+        submitted = st.form_submit_button("Check")
+
+    # Dataframe
+    data = {
+        'Holding_Policy_Duration': Holding_Policy_Duration,
+        'Holding_Policy_Type': Holding_Policy_Type,
+        'Reco_Policy_Cat': Reco_Policy_Cat,
+    }
+    df = pd.DataFrame([data])
+
+    # display dataframe of inputted data
+    st.dataframe(df)
+
+    # show result
+    if submitted:
+        result = model.predict(df)
+        if result == 1:
+            st.write('Lead will likely become actual customer')
+        else:
+            st.write('Lead will not likely become actual customer')
+
+if __name__=='__main__':
+    run()

requirements.txt

@@ -0,0 +1,5 @@
+pandas
+seaborn
+matplotlib
+pickleshare
+scikit-learn==1.3.0