Delete Хохлов_АА_T-7.ipynb_

Browse files

Files changed (1) hide show

Хохлов_АА_T-7.ipynb_ +0 -1

Хохлов_АА_T-7.ipynb_ DELETED Viewed

@@ -1 +0,0 @@

- {"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"provenance":[{"file_id":"1awAjxnallgqG4aWIoFm0adG955TQ4nyc","timestamp":1686721751642}]},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"}},"cells":[{"cell_type":"markdown","source":["Задание 7. Хохлов Александр Александрович 47 группа"],"metadata":{"id":"PjsWcR4OZ--n"}},{"cell_type":"code","source":["import numpy as np\n","from tensorflow import keras\n","from sklearn.model_selection import train_test_split\n","import matplotlib.pyplot as plt"],"metadata":{"id":"oeWkKW7UUQId"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Загрузка датасета Fashion MNIST\n","(X_train_full, y_train_full), (X_test, y_test) = keras.datasets.fashion_mnist.load_data()\n"],"metadata":{"id":"HFSeW-NIUSEF"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Нормализация данных\n","X_train_full = X_train_full.astype('float32') / 255.\n","X_test = X_test.astype('float32') / 255."],"metadata":{"id":"-3TRgkRCUSPU"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Разбиение на тренировочную, валидационную и тестовую части\n","X_train, X_val = train_test_split(X_train_full, test_size=0.2, random_state=42)\n","X_val, X_test = train_test_split(X_val, test_size=0.5, random_state=42)"],"metadata":{"id":"22owMkpMUUz4"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Создание нейросети\n","input_shape = X_train.shape[1:]\n","latent_dim = 50\n","autoencoder = keras.models.Sequential([\n"," keras.layers.Flatten(input_shape=input_shape),\n"," keras.layers.Dense(256, activation='relu'),\n"," keras.layers.Dense(128, activation='relu'),\n"," keras.layers.Dense(latent_dim, activation='relu', name='latent_layer'),\n"," keras.layers.Dense(128, activation='relu'),\n"," keras.layers.Dense(256, activation='relu'),\n"," keras.layers.Dense(np.prod(input_shape), activation='sigmoid'),\n"," keras.layers.Reshape(input_shape)\n","])\n"],"metadata":{"id":"Kr9q-sr8UXpN"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":[],"metadata":{"id":"goOHCrxmVDJG"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Компиляция и обучение.# binary_crossentropy может и не нужна, но тема рабочая, менять не буду.\n","autoencoder.compile(loss='binary_crossentropy', optimizer='adam',metrics=['accuracy'])\n","history = autoencoder.fit(X_train, X_train,\n"," epochs=50,\n"," batch_size=128,\n"," validation_data=(X_val, X_val))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"mAs8pBfKUd81","executionInfo":{"status":"ok","timestamp":1687273971835,"user_tz":-180,"elapsed":443800,"user":{"displayName":"Александр Хохлов","userId":"01141057562125644828"}},"outputId":"2ae96717-790c-4a76-9d47-b3a34275db10"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/50\n","375/375 [==============================] - 10s 23ms/step - loss: 0.2711 - accuracy: 0.2160 - val_loss: 0.2712 - val_accuracy: 0.2159\n","Epoch 2/50\n","375/375 [==============================] - 13s 34ms/step - loss: 0.2699 - accuracy: 0.2205 - val_loss: 0.2708 - val_accuracy: 0.2230\n","Epoch 3/50\n","375/375 [==============================] - 9s 24ms/step - loss: 0.2696 - accuracy: 0.2226 - val_loss: 0.2701 - val_accuracy: 0.2280\n","Epoch 4/50\n","375/375 [==============================] - 9s 23ms/step - loss: 0.2692 - accuracy: 0.2262 - val_loss: 0.2698 - val_accuracy: 0.2197\n","Epoch 5/50\n","375/375 [==============================] - 10s 26ms/step - loss: 0.2689 - accuracy: 0.2273 - val_loss: 0.2695 - val_accuracy: 0.2261\n","Epoch 6/50\n","375/375 [==============================] - 11s 28ms/step - loss: 0.2685 - accuracy: 0.2305 - val_loss: 0.2695 - val_accuracy: 0.2320\n","Epoch 7/50\n","375/375 [==============================] - 8s 20ms/step - loss: 0.2682 - accuracy: 0.2329 - val_loss: 0.2689 - val_accuracy: 0.2292\n","Epoch 8/50\n","375/375 [==============================] - 9s 23ms/step - loss: 0.2679 - accuracy: 0.2343 - val_loss: 0.2686 - val_accuracy: 0.2350\n","Epoch 9/50\n","375/375 [==============================] - 8s 21ms/step - loss: 0.2676 - accuracy: 0.2360 - val_loss: 0.2684 - val_accuracy: 0.2352\n","Epoch 10/50\n","375/375 [==============================] - 8s 23ms/step - loss: 0.2674 - accuracy: 0.2381 - val_loss: 0.2683 - val_accuracy: 0.2321\n","Epoch 11/50\n","375/375 [==============================] - 9s 23ms/step - loss: 0.2671 - accuracy: 0.2390 - val_loss: 0.2680 - val_accuracy: 0.2372\n","Epoch 12/50\n","375/375 [==============================] - 7s 19ms/step - loss: 0.2669 - accuracy: 0.2409 - val_loss: 0.2678 - val_accuracy: 0.2354\n","Epoch 13/50\n","375/375 [==============================] - 9s 23ms/step - loss: 0.2668 - accuracy: 0.2420 - val_loss: 0.2679 - val_accuracy: 0.2388\n","Epoch 14/50\n","375/375 [==============================] - 8s 20ms/step - loss: 0.2666 - accuracy: 0.2437 - val_loss: 0.2675 - val_accuracy: 0.2437\n","Epoch 15/50\n","375/375 [==============================] - 8s 20ms/step - loss: 0.2664 - accuracy: 0.2447 - val_loss: 0.2672 - val_accuracy: 0.2406\n","Epoch 16/50\n","375/375 [==============================] - 9s 23ms/step - loss: 0.2662 - accuracy: 0.2460 - val_loss: 0.2678 - val_accuracy: 0.2441\n","Epoch 17/50\n","375/375 [==============================] - 7s 20ms/step - loss: 0.2661 - accuracy: 0.2475 - val_loss: 0.2670 - val_accuracy: 0.2467\n","Epoch 18/50\n","375/375 [==============================] - 9s 23ms/step - loss: 0.2659 - accuracy: 0.2483 - val_loss: 0.2672 - val_accuracy: 0.2456\n","Epoch 19/50\n","375/375 [==============================] - 8s 22ms/step - loss: 0.2658 - accuracy: 0.2494 - val_loss: 0.2668 - val_accuracy: 0.2457\n","Epoch 20/50\n","375/375 [==============================] - 7s 20ms/step - loss: 0.2656 - accuracy: 0.2504 - val_loss: 0.2667 - val_accuracy: 0.2491\n","Epoch 21/50\n","375/375 [==============================] - 8s 22ms/step - loss: 0.2655 - accuracy: 0.2507 - val_loss: 0.2667 - val_accuracy: 0.2424\n","Epoch 22/50\n","375/375 [==============================] - 7s 19ms/step - loss: 0.2655 - accuracy: 0.2509 - val_loss: 0.2667 - val_accuracy: 0.2469\n","Epoch 23/50\n","375/375 [==============================] - 8s 23ms/step - loss: 0.2653 - accuracy: 0.2523 - val_loss: 0.2664 - val_accuracy: 0.2465\n","Epoch 24/50\n","375/375 [==============================] - 8s 22ms/step - loss: 0.2653 - accuracy: 0.2533 - val_loss: 0.2662 - val_accuracy: 0.2478\n","Epoch 25/50\n","375/375 [==============================] - 7s 20ms/step - loss: 0.2650 - accuracy: 0.2539 - val_loss: 0.2661 - val_accuracy: 0.2436\n","Epoch 26/50\n","375/375 [==============================] - 8s 23ms/step - loss: 0.2650 - accuracy: 0.2543 - val_loss: 0.2661 - val_accuracy: 0.2463\n","Epoch 27/50\n","375/375 [==============================] - 7s 19ms/step - loss: 0.2648 - accuracy: 0.2545 - val_loss: 0.2664 - val_accuracy: 0.2483\n","Epoch 28/50\n","375/375 [==============================] - 8s 22ms/step - loss: 0.2648 - accuracy: 0.2550 - val_loss: 0.2659 - val_accuracy: 0.2551\n","Epoch 29/50\n","375/375 [==============================] - 8s 20ms/step - loss: 0.2647 - accuracy: 0.2555 - val_loss: 0.2662 - val_accuracy: 0.2557\n","Epoch 30/50\n","375/375 [==============================] - 7s 20ms/step - loss: 0.2646 - accuracy: 0.2565 - val_loss: 0.2660 - val_accuracy: 0.2470\n","Epoch 31/50\n","375/375 [==============================] - 8s 22ms/step - loss: 0.2645 - accuracy: 0.2574 - val_loss: 0.2656 - val_accuracy: 0.2528\n","Epoch 32/50\n","375/375 [==============================] - 7s 20ms/step - loss: 0.2645 - accuracy: 0.2573 - val_loss: 0.2659 - val_accuracy: 0.2521\n","Epoch 33/50\n","375/375 [==============================] - 9s 23ms/step - loss: 0.2644 - accuracy: 0.2588 - val_loss: 0.2654 - val_accuracy: 0.2537\n","Epoch 34/50\n","375/375 [==============================] - 8s 21ms/step - loss: 0.2643 - accuracy: 0.2583 - val_loss: 0.2655 - val_accuracy: 0.2515\n","Epoch 35/50\n","375/375 [==============================] - 8s 20ms/step - loss: 0.2642 - accuracy: 0.2582 - val_loss: 0.2655 - val_accuracy: 0.2538\n","Epoch 36/50\n","375/375 [==============================] - 9s 23ms/step - loss: 0.2642 - accuracy: 0.2595 - val_loss: 0.2658 - val_accuracy: 0.2468\n","Epoch 37/50\n","375/375 [==============================] - 7s 19ms/step - loss: 0.2641 - accuracy: 0.2604 - val_loss: 0.2654 - val_accuracy: 0.2556\n","Epoch 38/50\n","375/375 [==============================] - 8s 23ms/step - loss: 0.2641 - accuracy: 0.2601 - val_loss: 0.2651 - val_accuracy: 0.2550\n","Epoch 39/50\n","375/375 [==============================] - 9s 25ms/step - loss: 0.2640 - accuracy: 0.2604 - val_loss: 0.2652 - val_accuracy: 0.2504\n","Epoch 40/50\n","375/375 [==============================] - 8s 22ms/step - loss: 0.2640 - accuracy: 0.2608 - val_loss: 0.2655 - val_accuracy: 0.2548\n","Epoch 41/50\n","375/375 [==============================] - 10s 27ms/step - loss: 0.2639 - accuracy: 0.2617 - val_loss: 0.2655 - val_accuracy: 0.2579\n","Epoch 42/50\n","375/375 [==============================] - 9s 25ms/step - loss: 0.2638 - accuracy: 0.2615 - val_loss: 0.2653 - val_accuracy: 0.2535\n","Epoch 43/50\n","375/375 [==============================] - 8s 21ms/step - loss: 0.2638 - accuracy: 0.2620 - val_loss: 0.2650 - val_accuracy: 0.2557\n","Epoch 44/50\n","375/375 [==============================] - 9s 25ms/step - loss: 0.2637 - accuracy: 0.2628 - val_loss: 0.2651 - val_accuracy: 0.2546\n","Epoch 45/50\n","375/375 [==============================] - 10s 26ms/step - loss: 0.2637 - accuracy: 0.2629 - val_loss: 0.2654 - val_accuracy: 0.2518\n","Epoch 46/50\n","375/375 [==============================] - 8s 20ms/step - loss: 0.2636 - accuracy: 0.2632 - val_loss: 0.2649 - val_accuracy: 0.2554\n","Epoch 47/50\n","375/375 [==============================] - 9s 23ms/step - loss: 0.2636 - accuracy: 0.2629 - val_loss: 0.2647 - val_accuracy: 0.2598\n","Epoch 48/50\n","375/375 [==============================] - 8s 21ms/step - loss: 0.2636 - accuracy: 0.2632 - val_loss: 0.2648 - val_accuracy: 0.2536\n","Epoch 49/50\n","375/375 [==============================] - 8s 21ms/step - loss: 0.2635 - accuracy: 0.2640 - val_loss: 0.2650 - val_accuracy: 0.2592\n","Epoch 50/50\n","375/375 [==============================] - 9s 23ms/step - loss: 0.2635 - accuracy: 0.2644 - val_loss: 0.2650 - val_accuracy: 0.2635\n"]}]},{"cell_type":"code","source":["# Оценка ошибки и точности на тренировочной, валидационной и тестовой части\n","train_loss, train_acc = autoencoder.evaluate(X_train, X_train)\n","val_loss, val_acc = autoencoder.evaluate(X_val, X_val)\n","test_loss, test_acc = autoencoder.evaluate(X_test, X_test)\n","\n","print(f'Train Loss: {train_loss:.5f}, Train Accuracy: {train_acc:.5f}')\n","print(f'Val Loss: {val_loss:.5f}, Val Accuracy: {val_acc:.5f}')\n","print(f'Test Loss: {test_loss:.5f}, Test Accuracy: {test_acc:.5f}')"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"4VpES8sFW7C4","executionInfo":{"status":"ok","timestamp":1687274177977,"user_tz":-180,"elapsed":24067,"user":{"displayName":"Александр Хохлов","userId":"01141057562125644828"}},"outputId":"fbbb9297-0a6c-42da-9a22-ff488de61df7"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["1500/1500 [==============================] - 7s 5ms/step - loss: 0.2635 - accuracy: 0.2715\n","188/188 [==============================] - 1s 4ms/step - loss: 0.2650 - accuracy: 0.2635\n","188/188 [==============================] - 1s 7ms/step - loss: 0.2644 - accuracy: 0.2660\n","Train Loss: 0.26351, Train Accuracy: 0.27152\n","Val Loss: 0.26502, Val Accuracy: 0.26352\n","Test Loss: 0.26442, Test Accuracy: 0.26600\n"]}]},{"cell_type":"code","source":["# Визуализация результатов. Я лично доволен. Оно работает!!!\n","n = 7 # количество изображений для примера\n","decoded_imgs = autoencoder.predict(X_test[:n]) # Кодировка и декодировка тестовых изображений\n","plt.figure(figsize=(10, 4.5))\n","for i in range(n):\n"," # Оригинальное изображение\n"," ax = plt.subplot(2, n, i + 1)\n"," plt.imshow(X_test[i])\n"," plt.gray()\n"," ax.get_xaxis().set_visible(False)\n"," ax.get_yaxis().set_visible(False)\n","\n"," # Декодированное изображение\n"," ax = plt.subplot(2, n, i + 1 + n)\n"," plt.imshow(decoded_imgs[i])\n"," plt.gray()\n"," ax.get_xaxis().set_visible(False)\n"," ax.get_yaxis().set_visible(False)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":337},"id":"sk7AZyb-UgfV","executionInfo":{"status":"ok","timestamp":1687274285712,"user_tz":-180,"elapsed":811,"user":{"displayName":"Александр Хохлов","userId":"01141057562125644828"}},"outputId":"504d1790-8119-4c9a-e1ab-caa3572c1fc2"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["1/1 [==============================] - 0s 32ms/step\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 1000x450 with 14 Axes>"],"image/png":"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\n"},"metadata":{}}]}]}