neural-network/lenet-50/lenet50-mnist.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "LeNet50.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "UhGsD1ftrNof",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# libs import\n",
        "from keras.layers.convolutional import Convolution2D\n",
        "from keras.layers.convolutional import MaxPooling2D\n",
        "from keras.layers.core import Activation\n",
        "from keras.layers.core import Flatten\n",
        "from keras.layers.core import Dense\n",
        "from keras.models import Sequential\n",
        "from keras import models, layers\n",
        "from keras.optimizers import SGD\n",
        "from keras.utils import np_utils\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.datasets import fetch_openml\n",
        "from tensorflow.python.util import deprecation\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import keras"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W4TwTyJ0wLbd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# fetching dataset\n",
        "dataset = fetch_openml('mnist_784')\n",
        "\n",
        "# suppress warning\n",
        "deprecation._PRINT_DEPRECATION_WARNINGS = False"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "K_b5bg9n2pld",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# reshaping\n",
        "data = dataset.data.reshape((dataset.data.shape[0], 28, 28))\n",
        "\n",
        "# reshaping + add new axis\n",
        "data = data[:, :, :, np.newaxis]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NN6eZCox10n4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# skalakan dari range 0-255 menjadi range 0-1\n",
        "scaled_data = data / 255.0"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vbNP7zfmxKDG",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# split training & test\n",
        "(x_train, x_test, y_train, y_test) = train_test_split(\n",
        "    scaled_data,\n",
        "    dataset.target.astype(\"float32\"), \n",
        "    test_size = 0.20)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HQeVRhS44sw1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# transformasi target training dengan one-hot encoding\n",
        "y_train = np_utils.to_categorical(y_train, 10)\n",
        "\n",
        "# transformasi target testing dengan one-hot encoding\n",
        "y_test = np_utils.to_categorical(y_test, 10)\n",
        "\n",
        "# One hot encoding is a process by which categorical variables are \n",
        "# converted into a form that could be provided to ML algorithms to \n",
        "# do a better job in prediction."
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qqphThjnxvLY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# instantiasikan model sequential kosong\n",
        "model = Sequential()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yBG2AymYyWPz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import warnings\n",
        "warnings.filterwarnings('ignore')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BDiTGS3M3kl2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# C1 Convolutional Layer\n",
        "model.add(layers.Conv2D(6, kernel_size=(5, 5), strides=(1, 1), activation=\"tanh\", input_shape=(28,28,1), padding=\"same\"))\n",
        "\n",
        "# S2 Pooling/Subsampling Layer\n",
        "model.add(layers.AveragePooling2D(pool_size=(2, 2), strides=(1, 1), padding=\"valid\"))\n",
        "\n",
        "# C3 Convolutional Layer\n",
        "model.add(layers.Conv2D(16, kernel_size=(5, 5), strides=(1, 1), activation=\"tanh\", padding=\"valid\"))\n",
        "\n",
        "# S4 Pooling/Subsampling Layer\n",
        "model.add(layers.AveragePooling2D(pool_size=(2, 2), strides=(2, 2), padding=\"valid\"))\n",
        "\n",
        "# C5 Fully Connected Convolutional Layer\n",
        "model.add(layers.Conv2D(120, kernel_size=(5, 5), strides=(1, 1), activation=\"tanh\", padding=\"valid\"))\n",
        "\n",
        "#Flatten the CNN output so that we can connect it with fully connected layers\n",
        "model.add(layers.Flatten())\n",
        "\n",
        "# FC6 Fully Connected Layer\n",
        "model.add(layers.Dense(84, activation=\"tanh\"))\n",
        "\n",
        "#Output Layer (softmax activation)\n",
        "model.add(layers.Dense(10, activation=\"softmax\"))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eeH-A8u0x87_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# model compiling\n",
        "model.compile(loss=keras.losses.categorical_crossentropy, optimizer=\"SGD\", metrics=[\"accuracy\"])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FF3ahqPQzS4p",
        "colab_type": "code",
        "outputId": "8a9fb142-f64c-447a-8082-57c0a416f2cc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "# epoch 50\n",
        "hist = model.fit(x=x_train,y=y_train, epochs=50, batch_size=128, validation_data=(x_test, y_test), verbose=1)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 56000 samples, validate on 14000 samples\n",
            "Epoch 1/50\n",
            "56000/56000 [==============================] - 3s 55us/step - loss: 0.6889 - acc: 0.8356 - val_loss: 0.3607 - val_acc: 0.8958\n",
            "Epoch 2/50\n",
            "56000/56000 [==============================] - 3s 52us/step - loss: 0.3135 - acc: 0.9120 - val_loss: 0.2845 - val_acc: 0.9203\n",
            "Epoch 3/50\n",
            "56000/56000 [==============================] - 3s 50us/step - loss: 0.2580 - acc: 0.9272 - val_loss: 0.2450 - val_acc: 0.9289\n",
            "Epoch 4/50\n",
            "56000/56000 [==============================] - 3s 51us/step - loss: 0.2239 - acc: 0.9356 - val_loss: 0.2168 - val_acc: 0.9379\n",
            "Epoch 5/50\n",
            "56000/56000 [==============================] - 3s 49us/step - loss: 0.1997 - acc: 0.9431 - val_loss: 0.1978 - val_acc: 0.9432\n",
            "Epoch 6/50\n",
            "56000/56000 [==============================] - 3s 53us/step - loss: 0.1810 - acc: 0.9484 - val_loss: 0.1826 - val_acc: 0.9472\n",
            "Epoch 7/50\n",
            "56000/56000 [==============================] - 3s 51us/step - loss: 0.1654 - acc: 0.9532 - val_loss: 0.1710 - val_acc: 0.9515\n",
            "Epoch 8/50\n",
            "56000/56000 [==============================] - 3s 50us/step - loss: 0.1525 - acc: 0.9569 - val_loss: 0.1571 - val_acc: 0.9547\n",
            "Epoch 9/50\n",
            "56000/56000 [==============================] - 3s 49us/step - loss: 0.1412 - acc: 0.9600 - val_loss: 0.1481 - val_acc: 0.9571\n",
            "Epoch 10/50\n",
            "56000/56000 [==============================] - 3s 49us/step - loss: 0.1316 - acc: 0.9630 - val_loss: 0.1387 - val_acc: 0.9596\n",
            "Epoch 11/50\n",
            "56000/56000 [==============================] - 3s 53us/step - loss: 0.1230 - acc: 0.9653 - val_loss: 0.1305 - val_acc: 0.9621\n",
            "Epoch 12/50\n",
            "56000/56000 [==============================] - 3s 49us/step - loss: 0.1151 - acc: 0.9683 - val_loss: 0.1241 - val_acc: 0.9630\n",
            "Epoch 13/50\n",
            "56000/56000 [==============================] - 3s 53us/step - loss: 0.1085 - acc: 0.9699 - val_loss: 0.1179 - val_acc: 0.9656\n",
            "Epoch 14/50\n",
            "56000/56000 [==============================] - 3s 52us/step - loss: 0.1023 - acc: 0.9715 - val_loss: 0.1131 - val_acc: 0.9666\n",
            "Epoch 15/50\n",
            "56000/56000 [==============================] - 3s 53us/step - loss: 0.0969 - acc: 0.9729 - val_loss: 0.1081 - val_acc: 0.9687\n",
            "Epoch 16/50\n",
            "56000/56000 [==============================] - 3s 51us/step - loss: 0.0916 - acc: 0.9747 - val_loss: 0.1027 - val_acc: 0.9699\n",
            "Epoch 17/50\n",
            "56000/56000 [==============================] - 3s 50us/step - loss: 0.0870 - acc: 0.9760 - val_loss: 0.0989 - val_acc: 0.9715\n",
            "Epoch 18/50\n",
            "56000/56000 [==============================] - 3s 51us/step - loss: 0.0830 - acc: 0.9773 - val_loss: 0.0952 - val_acc: 0.9724\n",
            "Epoch 19/50\n",
            "56000/56000 [==============================] - 3s 48us/step - loss: 0.0789 - acc: 0.9782 - val_loss: 0.0919 - val_acc: 0.9731\n",
            "Epoch 20/50\n",
            "56000/56000 [==============================] - 3s 51us/step - loss: 0.0754 - acc: 0.9795 - val_loss: 0.0886 - val_acc: 0.9744\n",
            "Epoch 21/50\n",
            "56000/56000 [==============================] - 3s 52us/step - loss: 0.0721 - acc: 0.9804 - val_loss: 0.0859 - val_acc: 0.9751\n",
            "Epoch 22/50\n",
            "56000/56000 [==============================] - 3s 51us/step - loss: 0.0689 - acc: 0.9814 - val_loss: 0.0831 - val_acc: 0.9762\n",
            "Epoch 23/50\n",
            "56000/56000 [==============================] - 3s 52us/step - loss: 0.0662 - acc: 0.9824 - val_loss: 0.0808 - val_acc: 0.9761\n",
            "Epoch 24/50\n",
            "56000/56000 [==============================] - 3s 49us/step - loss: 0.0635 - acc: 0.9830 - val_loss: 0.0799 - val_acc: 0.9769\n",
            "Epoch 25/50\n",
            "56000/56000 [==============================] - 3s 49us/step - loss: 0.0612 - acc: 0.9837 - val_loss: 0.0753 - val_acc: 0.9777\n",
            "Epoch 26/50\n",
            "56000/56000 [==============================] - 3s 50us/step - loss: 0.0589 - acc: 0.9840 - val_loss: 0.0740 - val_acc: 0.9782\n",
            "Epoch 27/50\n",
            "56000/56000 [==============================] - 3s 47us/step - loss: 0.0566 - acc: 0.9852 - val_loss: 0.0714 - val_acc: 0.9791\n",
            "Epoch 28/50\n",
            "56000/56000 [==============================] - 3s 49us/step - loss: 0.0547 - acc: 0.9851 - val_loss: 0.0700 - val_acc: 0.9789\n",
            "Epoch 29/50\n",
            "56000/56000 [==============================] - 3s 49us/step - loss: 0.0526 - acc: 0.9861 - val_loss: 0.0683 - val_acc: 0.9799\n",
            "Epoch 30/50\n",
            "56000/56000 [==============================] - 3s 50us/step - loss: 0.0511 - acc: 0.9865 - val_loss: 0.0687 - val_acc: 0.9794\n",
            "Epoch 31/50\n",
            "56000/56000 [==============================] - 3s 49us/step - loss: 0.0494 - acc: 0.9869 - val_loss: 0.0661 - val_acc: 0.9802\n",
            "Epoch 32/50\n",
            "56000/56000 [==============================] - 3s 51us/step - loss: 0.0477 - acc: 0.9874 - val_loss: 0.0646 - val_acc: 0.9810\n",
            "Epoch 33/50\n",
            "56000/56000 [==============================] - 3s 48us/step - loss: 0.0462 - acc: 0.9877 - val_loss: 0.0641 - val_acc: 0.9809\n",
            "Epoch 34/50\n",
            "56000/56000 [==============================] - 3s 53us/step - loss: 0.0448 - acc: 0.9883 - val_loss: 0.0621 - val_acc: 0.9813\n",
            "Epoch 35/50\n",
            "56000/56000 [==============================] - 3s 51us/step - loss: 0.0433 - acc: 0.9886 - val_loss: 0.0610 - val_acc: 0.9816\n",
            "Epoch 36/50\n",
            "56000/56000 [==============================] - 3s 52us/step - loss: 0.0420 - acc: 0.9890 - val_loss: 0.0601 - val_acc: 0.9819\n",
            "Epoch 37/50\n",
            "56000/56000 [==============================] - 3s 52us/step - loss: 0.0408 - acc: 0.9895 - val_loss: 0.0590 - val_acc: 0.9823\n",
            "Epoch 38/50\n",
            "56000/56000 [==============================] - 3s 50us/step - loss: 0.0397 - acc: 0.9897 - val_loss: 0.0576 - val_acc: 0.9819\n",
            "Epoch 39/50\n",
            "56000/56000 [==============================] - 3s 47us/step - loss: 0.0385 - acc: 0.9900 - val_loss: 0.0569 - val_acc: 0.9824\n",
            "Epoch 40/50\n",
            "56000/56000 [==============================] - 3s 51us/step - loss: 0.0374 - acc: 0.9905 - val_loss: 0.0566 - val_acc: 0.9823\n",
            "Epoch 41/50\n",
            "56000/56000 [==============================] - 3s 52us/step - loss: 0.0365 - acc: 0.9906 - val_loss: 0.0569 - val_acc: 0.9834\n",
            "Epoch 42/50\n",
            "56000/56000 [==============================] - 3s 48us/step - loss: 0.0354 - acc: 0.9910 - val_loss: 0.0582 - val_acc: 0.9820\n",
            "Epoch 43/50\n",
            "56000/56000 [==============================] - 3s 51us/step - loss: 0.0346 - acc: 0.9911 - val_loss: 0.0538 - val_acc: 0.9829\n",
            "Epoch 44/50\n",
            "56000/56000 [==============================] - 3s 50us/step - loss: 0.0335 - acc: 0.9916 - val_loss: 0.0536 - val_acc: 0.9836\n",
            "Epoch 45/50\n",
            "56000/56000 [==============================] - 3s 51us/step - loss: 0.0327 - acc: 0.9918 - val_loss: 0.0535 - val_acc: 0.9836\n",
            "Epoch 46/50\n",
            "56000/56000 [==============================] - 3s 49us/step - loss: 0.0320 - acc: 0.9921 - val_loss: 0.0517 - val_acc: 0.9840\n",
            "Epoch 47/50\n",
            "56000/56000 [==============================] - 3s 49us/step - loss: 0.0311 - acc: 0.9920 - val_loss: 0.0514 - val_acc: 0.9842\n",
            "Epoch 48/50\n",
            "56000/56000 [==============================] - 3s 52us/step - loss: 0.0303 - acc: 0.9925 - val_loss: 0.0508 - val_acc: 0.9840\n",
            "Epoch 49/50\n",
            "56000/56000 [==============================] - 3s 50us/step - loss: 0.0297 - acc: 0.9927 - val_loss: 0.0502 - val_acc: 0.9842\n",
            "Epoch 50/50\n",
            "56000/56000 [==============================] - 3s 53us/step - loss: 0.0288 - acc: 0.9931 - val_loss: 0.0504 - val_acc: 0.9842\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yyVkdZKl1Upl",
        "colab_type": "code",
        "outputId": "3d16f759-da83-43cf-be52-26d881380d8d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "test_score = model.evaluate(x_test, y_test)\n",
        "print(\"Test loss {:.4f}, accuracy {:.2f}%\".format(test_score[0], test_score[1] * 100))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "14000/14000 [==============================] - 1s 52us/step\n",
            "Test loss 0.0504, accuracy 98.42%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZMuwlHrk3c9P",
        "colab_type": "code",
        "outputId": "2b3b3473-3f02-4517-db28-9fdcf1b87281",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 312
        }
      },
      "source": [
        "#plotting\n",
        "f, ax = plt.subplots()\n",
        "ax.plot([None] + hist.history[\"acc\"], \"o-\")\n",
        "ax.plot([None] + hist.history[\"val_acc\"], \"x-\")\n",
        "\n",
        "ax.legend([\"Train acc\", \"Validation acc\"], loc = 0)\n",
        "ax.set_title(\"Training/Validation acc per Epoch\")\n",
        "ax.set_xlabel(\"Epoch\")\n",
        "ax.set_ylabel(\"acc\") "
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'acc')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Qd1nDe5D5YiI",
        "colab_type": "code",
        "outputId": "8b0a6439-6d69-44ee-c938-407a66a1e0c4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 312
        }
      },
      "source": [
        "#plotting\n",
        "f, ax = plt.subplots()\n",
        "ax.plot([None] + hist.history[\"loss\"])\n",
        "ax.plot([None] + hist.history[\"val_loss\"])\n",
        "\n",
        "ax.legend([\"Train acc\", \"Validation acc\"], loc = 0)\n",
        "ax.set_title(\"Training/Validation acc per Epoch\")\n",
        "ax.set_xlabel(\"Epoch\")\n",
        "ax.set_ylabel(\"acc\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'acc')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}