Deep Learning Base Model Cats-Dogs Datasets.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Assignment15_RickyGunawan_Base.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "sCn4y6io1E07"
      },
      "source": [
        "zip_path = '/content/drive/My\\ Drive/BCML/datasets/DeepLearning/cats-dogs.zip'\n",
        "\n",
        "!cp {zip_path} /content/\n",
        "\n",
        "!cd /content/\n",
        "\n",
        "!unzip -q /content/cats-dogs.zip -d /content\n",
        "\n",
        "!rm /content/cats-dogs.zip"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ek83tzRrIy7Y"
      },
      "source": [
        "def plot_loss(history):\n",
        "    plt.plot(history.history['loss'], label='loss')\n",
        "    plt.plot(history.history['val_loss'], label='val_loss')\n",
        "    plt.plot(history.history['val_accuracy'], label='val_accuracy')\n",
        "    plt.xlabel('Epoch')\n",
        "    plt.ylabel('Loss')\n",
        "    plt.legend()\n",
        "    plt.grid(True)"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IXyxQWL7v12W"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import os\n",
        "\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.utils import to_categorical\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Conv2D\n",
        "from tensorflow.keras.layers import MaxPooling2D\n",
        "from tensorflow.keras.layers import Dense\n",
        "from tensorflow.keras.layers import Flatten\n",
        "from tensorflow.keras.optimizers import SGD\n",
        "from tensorflow.keras.layers import BatchNormalization"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ObPuYkN8ql-O"
      },
      "source": [
        "# **Base Model**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SlSM8yCJwSXQ",
        "outputId": "d54717b4-f68c-487d-dbe9-42c2ea2cde6d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 323
        }
      },
      "source": [
        "model = Sequential()\n",
        "model.add(Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same', input_shape=(200, 200, 3)))\n",
        "model.add(MaxPooling2D((2, 2)))\n",
        "\n",
        "model.add(Flatten())\n",
        "\n",
        "model.add(Dense(128, activation='relu', kernel_initializer='he_uniform'))\n",
        "model.add(Dense(1, activation='sigmoid'))\n",
        "\n",
        "opt = SGD(lr=0.001, momentum=0.9)\n",
        "\n",
        "model.compile(optimizer=opt, loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 200, 200, 32)      896       \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 100, 100, 32)      0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 320000)            0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 128)               40960128  \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 1)                 129       \n",
            "=================================================================\n",
            "Total params: 40,961,153\n",
            "Trainable params: 40,961,153\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3OUmunEpzExb",
        "outputId": "7971f02b-4ff9-4c22-8569-12475f4409fe",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "dataset_dir = '/content/'\n",
        "\n",
        "datagen = ImageDataGenerator(rescale=1.0/255.0)\n",
        "\n",
        "train_iterator = datagen.flow_from_directory(os.path.join(dataset_dir, 'train'), class_mode='binary', batch_size=128, target_size=(200, 200))\n",
        "test_iterator = datagen.flow_from_directory(os.path.join(dataset_dir, 'test'), class_mode='binary', batch_size=128, target_size=(200, 200))"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 12749 images belonging to 2 classes.\n",
            "Found 2252 images belonging to 2 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SumVfASyzyzs",
        "outputId": "2ce731f0-8355-4cf6-94e7-a6e801fa8f7d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "history=model.fit(train_iterator, steps_per_epoch=len(train_iterator), validation_data=test_iterator, validation_steps=len(test_iterator), epochs=75)"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/75\n",
            "100/100 [==============================] - 47s 466ms/step - loss: 0.8080 - accuracy: 0.5449 - val_loss: 0.6788 - val_accuracy: 0.5524\n",
            "Epoch 2/75\n",
            "100/100 [==============================] - 46s 462ms/step - loss: 0.6677 - accuracy: 0.5924 - val_loss: 0.6499 - val_accuracy: 0.6190\n",
            "Epoch 3/75\n",
            "100/100 [==============================] - 47s 469ms/step - loss: 0.6457 - accuracy: 0.6181 - val_loss: 0.6370 - val_accuracy: 0.6212\n",
            "Epoch 4/75\n",
            "100/100 [==============================] - 47s 466ms/step - loss: 0.6324 - accuracy: 0.6336 - val_loss: 0.6700 - val_accuracy: 0.5884\n",
            "Epoch 5/75\n",
            "100/100 [==============================] - 46s 463ms/step - loss: 0.6286 - accuracy: 0.6393 - val_loss: 0.6220 - val_accuracy: 0.6430\n",
            "Epoch 6/75\n",
            "100/100 [==============================] - 46s 461ms/step - loss: 0.6222 - accuracy: 0.6453 - val_loss: 0.6371 - val_accuracy: 0.6070\n",
            "Epoch 7/75\n",
            "100/100 [==============================] - 46s 459ms/step - loss: 0.6158 - accuracy: 0.6524 - val_loss: 0.6299 - val_accuracy: 0.6470\n",
            "Epoch 8/75\n",
            "100/100 [==============================] - 45s 453ms/step - loss: 0.6161 - accuracy: 0.6539 - val_loss: 0.6144 - val_accuracy: 0.6594\n",
            "Epoch 9/75\n",
            "100/100 [==============================] - 45s 451ms/step - loss: 0.6066 - accuracy: 0.6625 - val_loss: 0.6389 - val_accuracy: 0.6306\n",
            "Epoch 10/75\n",
            "100/100 [==============================] - 45s 453ms/step - loss: 0.5922 - accuracy: 0.6783 - val_loss: 0.5989 - val_accuracy: 0.6665\n",
            "Epoch 11/75\n",
            "100/100 [==============================] - 46s 457ms/step - loss: 0.5948 - accuracy: 0.6709 - val_loss: 0.6125 - val_accuracy: 0.6585\n",
            "Epoch 12/75\n",
            "100/100 [==============================] - 46s 465ms/step - loss: 0.5745 - accuracy: 0.6972 - val_loss: 0.6065 - val_accuracy: 0.6639\n",
            "Epoch 13/75\n",
            "100/100 [==============================] - 46s 461ms/step - loss: 0.5842 - accuracy: 0.6869 - val_loss: 0.6444 - val_accuracy: 0.6190\n",
            "Epoch 14/75\n",
            "100/100 [==============================] - 46s 455ms/step - loss: 0.5877 - accuracy: 0.6816 - val_loss: 0.6290 - val_accuracy: 0.6345\n",
            "Epoch 15/75\n",
            "100/100 [==============================] - 46s 456ms/step - loss: 0.5538 - accuracy: 0.7136 - val_loss: 0.5884 - val_accuracy: 0.6816\n",
            "Epoch 16/75\n",
            "100/100 [==============================] - 46s 456ms/step - loss: 0.5495 - accuracy: 0.7188 - val_loss: 0.5859 - val_accuracy: 0.6776\n",
            "Epoch 17/75\n",
            "100/100 [==============================] - 46s 456ms/step - loss: 0.5630 - accuracy: 0.7072 - val_loss: 0.5928 - val_accuracy: 0.6758\n",
            "Epoch 18/75\n",
            "100/100 [==============================] - 46s 457ms/step - loss: 0.5326 - accuracy: 0.7307 - val_loss: 0.5862 - val_accuracy: 0.6909\n",
            "Epoch 19/75\n",
            "100/100 [==============================] - 46s 458ms/step - loss: 0.5258 - accuracy: 0.7393 - val_loss: 0.6197 - val_accuracy: 0.6710\n",
            "Epoch 20/75\n",
            "100/100 [==============================] - 46s 455ms/step - loss: 0.5239 - accuracy: 0.7367 - val_loss: 0.5719 - val_accuracy: 0.6998\n",
            "Epoch 21/75\n",
            "100/100 [==============================] - 45s 454ms/step - loss: 0.5072 - accuracy: 0.7492 - val_loss: 0.5868 - val_accuracy: 0.6887\n",
            "Epoch 22/75\n",
            "100/100 [==============================] - 46s 458ms/step - loss: 0.5022 - accuracy: 0.7525 - val_loss: 0.5753 - val_accuracy: 0.7060\n",
            "Epoch 23/75\n",
            "100/100 [==============================] - 46s 455ms/step - loss: 0.4779 - accuracy: 0.7732 - val_loss: 0.5878 - val_accuracy: 0.6936\n",
            "Epoch 24/75\n",
            "100/100 [==============================] - 45s 453ms/step - loss: 0.4752 - accuracy: 0.7779 - val_loss: 0.7565 - val_accuracy: 0.5653\n",
            "Epoch 25/75\n",
            "100/100 [==============================] - 45s 455ms/step - loss: 0.5078 - accuracy: 0.7472 - val_loss: 0.5577 - val_accuracy: 0.7123\n",
            "Epoch 26/75\n",
            "100/100 [==============================] - 45s 452ms/step - loss: 0.4527 - accuracy: 0.7939 - val_loss: 0.6550 - val_accuracy: 0.6741\n",
            "Epoch 27/75\n",
            "100/100 [==============================] - 45s 451ms/step - loss: 0.4449 - accuracy: 0.7953 - val_loss: 0.6079 - val_accuracy: 0.6758\n",
            "Epoch 28/75\n",
            "100/100 [==============================] - 45s 450ms/step - loss: 0.4234 - accuracy: 0.8130 - val_loss: 0.5552 - val_accuracy: 0.7163\n",
            "Epoch 29/75\n",
            "100/100 [==============================] - 45s 454ms/step - loss: 0.3899 - accuracy: 0.8338 - val_loss: 0.5622 - val_accuracy: 0.7238\n",
            "Epoch 30/75\n",
            "100/100 [==============================] - 45s 451ms/step - loss: 0.3758 - accuracy: 0.8409 - val_loss: 0.5601 - val_accuracy: 0.7167\n",
            "Epoch 31/75\n",
            "100/100 [==============================] - 45s 452ms/step - loss: 0.3430 - accuracy: 0.8616 - val_loss: 0.5952 - val_accuracy: 0.7083\n",
            "Epoch 32/75\n",
            "100/100 [==============================] - 45s 455ms/step - loss: 0.3263 - accuracy: 0.8692 - val_loss: 0.5818 - val_accuracy: 0.7207\n",
            "Epoch 33/75\n",
            "100/100 [==============================] - 45s 452ms/step - loss: 0.3105 - accuracy: 0.8761 - val_loss: 0.5657 - val_accuracy: 0.7296\n",
            "Epoch 34/75\n",
            "100/100 [==============================] - 45s 451ms/step - loss: 0.3007 - accuracy: 0.8791 - val_loss: 0.5739 - val_accuracy: 0.7282\n",
            "Epoch 35/75\n",
            "100/100 [==============================] - 45s 449ms/step - loss: 0.2813 - accuracy: 0.8913 - val_loss: 0.6312 - val_accuracy: 0.7100\n",
            "Epoch 36/75\n",
            "100/100 [==============================] - 45s 450ms/step - loss: 0.2803 - accuracy: 0.8907 - val_loss: 0.5625 - val_accuracy: 0.7358\n",
            "Epoch 37/75\n",
            "100/100 [==============================] - 45s 448ms/step - loss: 0.2443 - accuracy: 0.9158 - val_loss: 0.5773 - val_accuracy: 0.7291\n",
            "Epoch 38/75\n",
            "100/100 [==============================] - 45s 450ms/step - loss: 0.2441 - accuracy: 0.9049 - val_loss: 0.5896 - val_accuracy: 0.7220\n",
            "Epoch 39/75\n",
            "100/100 [==============================] - 45s 453ms/step - loss: 0.2361 - accuracy: 0.9147 - val_loss: 0.5754 - val_accuracy: 0.7340\n",
            "Epoch 40/75\n",
            "100/100 [==============================] - 45s 452ms/step - loss: 0.2114 - accuracy: 0.9288 - val_loss: 0.5811 - val_accuracy: 0.7425\n",
            "Epoch 41/75\n",
            "100/100 [==============================] - 45s 450ms/step - loss: 0.1868 - accuracy: 0.9439 - val_loss: 0.6057 - val_accuracy: 0.7340\n",
            "Epoch 42/75\n",
            "100/100 [==============================] - 45s 447ms/step - loss: 0.1711 - accuracy: 0.9510 - val_loss: 0.6439 - val_accuracy: 0.7265\n",
            "Epoch 43/75\n",
            "100/100 [==============================] - 45s 453ms/step - loss: 0.1621 - accuracy: 0.9520 - val_loss: 0.6254 - val_accuracy: 0.7300\n",
            "Epoch 44/75\n",
            "100/100 [==============================] - 45s 451ms/step - loss: 0.1402 - accuracy: 0.9662 - val_loss: 0.6333 - val_accuracy: 0.7313\n",
            "Epoch 45/75\n",
            "100/100 [==============================] - 45s 452ms/step - loss: 0.1293 - accuracy: 0.9704 - val_loss: 0.7144 - val_accuracy: 0.7180\n",
            "Epoch 46/75\n",
            "100/100 [==============================] - 45s 452ms/step - loss: 0.1220 - accuracy: 0.9726 - val_loss: 0.7044 - val_accuracy: 0.7247\n",
            "Epoch 47/75\n",
            "100/100 [==============================] - 45s 450ms/step - loss: 0.1130 - accuracy: 0.9754 - val_loss: 0.8140 - val_accuracy: 0.7012\n",
            "Epoch 48/75\n",
            "100/100 [==============================] - 45s 450ms/step - loss: 0.1140 - accuracy: 0.9724 - val_loss: 0.6722 - val_accuracy: 0.7331\n",
            "Epoch 49/75\n",
            "100/100 [==============================] - 45s 448ms/step - loss: 0.0913 - accuracy: 0.9863 - val_loss: 0.7029 - val_accuracy: 0.7282\n",
            "Epoch 50/75\n",
            "100/100 [==============================] - 45s 451ms/step - loss: 0.0853 - accuracy: 0.9857 - val_loss: 0.7324 - val_accuracy: 0.7225\n",
            "Epoch 51/75\n",
            "100/100 [==============================] - 45s 450ms/step - loss: 0.0792 - accuracy: 0.9881 - val_loss: 0.7068 - val_accuracy: 0.7265\n",
            "Epoch 52/75\n",
            "100/100 [==============================] - 45s 450ms/step - loss: 0.0812 - accuracy: 0.9862 - val_loss: 0.7335 - val_accuracy: 0.7216\n",
            "Epoch 53/75\n",
            "100/100 [==============================] - 45s 449ms/step - loss: 0.0783 - accuracy: 0.9856 - val_loss: 0.7283 - val_accuracy: 0.7282\n",
            "Epoch 54/75\n",
            "100/100 [==============================] - 47s 465ms/step - loss: 0.0597 - accuracy: 0.9944 - val_loss: 0.7315 - val_accuracy: 0.7296\n",
            "Epoch 55/75\n",
            "100/100 [==============================] - 46s 458ms/step - loss: 0.0559 - accuracy: 0.9945 - val_loss: 0.7664 - val_accuracy: 0.7251\n",
            "Epoch 56/75\n",
            "100/100 [==============================] - 46s 457ms/step - loss: 0.0497 - accuracy: 0.9963 - val_loss: 0.7561 - val_accuracy: 0.7340\n",
            "Epoch 57/75\n",
            "100/100 [==============================] - 46s 456ms/step - loss: 0.0552 - accuracy: 0.9930 - val_loss: 0.7624 - val_accuracy: 0.7260\n",
            "Epoch 58/75\n",
            "100/100 [==============================] - 45s 453ms/step - loss: 0.0632 - accuracy: 0.9893 - val_loss: 0.7560 - val_accuracy: 0.7349\n",
            "Epoch 59/75\n",
            "100/100 [==============================] - 46s 455ms/step - loss: 0.0426 - accuracy: 0.9979 - val_loss: 0.7773 - val_accuracy: 0.7340\n",
            "Epoch 60/75\n",
            "100/100 [==============================] - 46s 459ms/step - loss: 0.0384 - accuracy: 0.9975 - val_loss: 0.7865 - val_accuracy: 0.7282\n",
            "Epoch 61/75\n",
            "100/100 [==============================] - 46s 461ms/step - loss: 0.0347 - accuracy: 0.9983 - val_loss: 0.8101 - val_accuracy: 0.7305\n",
            "Epoch 62/75\n",
            "100/100 [==============================] - 46s 461ms/step - loss: 0.0317 - accuracy: 0.9989 - val_loss: 0.8075 - val_accuracy: 0.7282\n",
            "Epoch 63/75\n",
            "100/100 [==============================] - 46s 458ms/step - loss: 0.0291 - accuracy: 0.9986 - val_loss: 0.8285 - val_accuracy: 0.7296\n",
            "Epoch 64/75\n",
            "100/100 [==============================] - 46s 460ms/step - loss: 0.0285 - accuracy: 0.9987 - val_loss: 0.8491 - val_accuracy: 0.7207\n",
            "Epoch 65/75\n",
            "100/100 [==============================] - 46s 461ms/step - loss: 0.0273 - accuracy: 0.9991 - val_loss: 0.8484 - val_accuracy: 0.7349\n",
            "Epoch 66/75\n",
            "100/100 [==============================] - 46s 465ms/step - loss: 0.0249 - accuracy: 0.9992 - val_loss: 0.8431 - val_accuracy: 0.7291\n",
            "Epoch 67/75\n",
            "100/100 [==============================] - 46s 462ms/step - loss: 0.0238 - accuracy: 0.9992 - val_loss: 0.8657 - val_accuracy: 0.7282\n",
            "Epoch 68/75\n",
            "100/100 [==============================] - 46s 460ms/step - loss: 0.0216 - accuracy: 0.9994 - val_loss: 0.8746 - val_accuracy: 0.7309\n",
            "Epoch 69/75\n",
            "100/100 [==============================] - 46s 461ms/step - loss: 0.0215 - accuracy: 0.9994 - val_loss: 0.8837 - val_accuracy: 0.7287\n",
            "Epoch 70/75\n",
            "100/100 [==============================] - 46s 455ms/step - loss: 0.0210 - accuracy: 0.9993 - val_loss: 0.8913 - val_accuracy: 0.7265\n",
            "Epoch 71/75\n",
            "100/100 [==============================] - 46s 457ms/step - loss: 0.0184 - accuracy: 0.9995 - val_loss: 0.9118 - val_accuracy: 0.7256\n",
            "Epoch 72/75\n",
            "100/100 [==============================] - 46s 455ms/step - loss: 0.0181 - accuracy: 0.9998 - val_loss: 0.8926 - val_accuracy: 0.7318\n",
            "Epoch 73/75\n",
            "100/100 [==============================] - 46s 456ms/step - loss: 0.0161 - accuracy: 0.9998 - val_loss: 0.9055 - val_accuracy: 0.7309\n",
            "Epoch 74/75\n",
            "100/100 [==============================] - 45s 451ms/step - loss: 0.0168 - accuracy: 0.9998 - val_loss: 0.9110 - val_accuracy: 0.7327\n",
            "Epoch 75/75\n",
            "100/100 [==============================] - 45s 449ms/step - loss: 0.0150 - accuracy: 0.9998 - val_loss: 0.9362 - val_accuracy: 0.7287\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vOgx5QygI0aP",
        "outputId": "cc19654c-2bda-4a07-853a-a4a7ed8bc156",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        }
      },
      "source": [
        "plot_loss(history)"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YvptLOT0qqYy"
      },
      "source": [
        "Kesimpulan ada pada file Assignment15_RickyGunawan_imageaug_filteradd"
      ]
    }
  ]
}