01_Simple_neural_network(mlp)_in_keras.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mUXNXqK7n0Tt"
      },
      "source": [
        "<center><div style=\"direction:rtl;font-family:B Lotus, B Nazanin, Tahoma\">به نام خدا</div></center>\n",
        "\n",
        "<h1><center><div style=\"direction:rtl;font-family:B Lotus, B Nazanin, Tahoma\">مقدمه‌ای بر شبکه‌های عصبی و چارچوب Keras (کراس)</div></center></h1>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jP3IhlAYn0Tw"
      },
      "source": [
        "## <div style=\"direction:rtl;text-align:right;font-family:B Lotus, B Nazanin, Tahoma\">لود کتابخانه‌های مورد استفاده</div>\n",
        "<div style=\"direction:rtl;text-align:right;font-family:Tahoma\">\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!wget https://raw.githubusercontent.com/Alireza-Akhavan/deeplearning-tensorflow2-notebooks/master/dataset.py\n",
        "!wget https://github.com/Alireza-Akhavan/deeplearning-tensorflow2-notebooks/raw/master/dataset/Data_hoda_full.mat -P dataset"
      ],
      "metadata": {
        "id": "TbJqL_8poKzA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "nbpresent": {
          "id": "de829a92-1fb6-44ad-a2c6-fc1001e1f6e1"
        },
        "id": "wRZGKJERn0Ty"
      },
      "outputs": [],
      "source": [
        "from tensorflow import keras\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense\n",
        "import numpy as np\n",
        "\n",
        "from dataset import load_hoda"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n6Fce1Mmn0T0"
      },
      "source": [
        "## <div style=\"direction:rtl;text-align:right;font-family:B Lotus, B Nazanin, Tahoma\">لود مجموعه داده (dataset)</div>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "75933Yn4n0T2"
      },
      "outputs": [],
      "source": [
        "x_train_original, y_train_original, x_test_original, y_test_original = load_hoda()"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_train_original[:3]"
      ],
      "metadata": {
        "id": "Ea33wtOhojtO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "TNqdNuDmn0T4"
      },
      "outputs": [],
      "source": [
        "# Preprocess input data for Keras.\n",
        "y_train = keras.utils.to_categorical(y_train_original, num_classes=10)\n",
        "y_test = keras.utils.to_categorical(y_test_original, num_classes=10)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_train[:3]"
      ],
      "metadata": {
        "id": "uoO2KjILomRf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HTMmpwKvn0T6"
      },
      "source": [
        "<div style=\"direction:rtl;text-align:right;font-family:Tahoma\">\n",
        "نرمال سازی مقادیر به مقدار بین 0 و 1\n",
        "</div>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Qq6rwPV3n0T7"
      },
      "outputs": [],
      "source": [
        "x_train = x_train_original.astype('float32')\n",
        "x_test = x_test_original.astype('float32')\n",
        "x_train /= 255\n",
        "x_test /= 255"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jnNoMbA4n0UA"
      },
      "source": [
        "## <div style=\"direction:rtl;text-align:right;font-family:B Lotus, B Nazanin, Tahoma\">تعریف معماری مدل (model architecture)</div>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "yXI8Qt9ln0UD"
      },
      "outputs": [],
      "source": [
        "model = Sequential()\n",
        "model.add(Dense(64, activation='relu', input_dim=25))\n",
        "model.add(Dense(10, activation='softmax'))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "PjjzCgUon0UE"
      },
      "outputs": [],
      "source": [
        "model.summary()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eiBtCFgsn0UG"
      },
      "source": [
        "## <div style=\"direction:rtl;text-align:right;font-family:B Lotus, B Nazanin, Tahoma\">Compile model</div>\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "LgsEJt-on0UH"
      },
      "outputs": [],
      "source": [
        "model.compile(loss='categorical_crossentropy',\n",
        "              optimizer='rmsprop',\n",
        "              metrics=['accuracy'])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "K90TipVyn0UI"
      },
      "source": [
        "## <div style=\"direction:rtl;text-align:right;font-family:B Lotus, B Nazanin, Tahoma\">آموش مدل با داده‌های آموزشی</div>\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "1000*0.8"
      ],
      "metadata": {
        "id": "fII6tnXYp9B4",
        "outputId": "67134075-f97b-48cb-bed5-3d7cfeccebee",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "800.0"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import math\n",
        "math.ceil((800/64))"
      ],
      "metadata": {
        "id": "j_5xOZUSqES1",
        "outputId": "2231f503-88e4-461a-fd3d-9a4c31ef5645",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "13"
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "pLBWKD3_n0UK",
        "outputId": "b7a3cf90-d1b1-4435-8eac-b2ae13396a51",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "13/13 [==============================] - 5s 19ms/step - loss: 2.2247 - accuracy: 0.1950 - val_loss: 2.0634 - val_accuracy: 0.3900\n",
            "Epoch 2/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 1.9780 - accuracy: 0.4125 - val_loss: 1.8480 - val_accuracy: 0.5950\n",
            "Epoch 3/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 1.7894 - accuracy: 0.5987 - val_loss: 1.6616 - val_accuracy: 0.7250\n",
            "Epoch 4/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 1.6171 - accuracy: 0.7113 - val_loss: 1.4928 - val_accuracy: 0.7850\n",
            "Epoch 5/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 1.4546 - accuracy: 0.7613 - val_loss: 1.3412 - val_accuracy: 0.8000\n",
            "Epoch 6/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 1.3066 - accuracy: 0.7800 - val_loss: 1.2045 - val_accuracy: 0.8050\n",
            "Epoch 7/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 1.1730 - accuracy: 0.8138 - val_loss: 1.0794 - val_accuracy: 0.8050\n",
            "Epoch 8/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 1.0531 - accuracy: 0.8150 - val_loss: 0.9778 - val_accuracy: 0.8250\n",
            "Epoch 9/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.9496 - accuracy: 0.8400 - val_loss: 0.8870 - val_accuracy: 0.8100\n",
            "Epoch 10/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.8590 - accuracy: 0.8487 - val_loss: 0.8096 - val_accuracy: 0.8250\n",
            "Epoch 11/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.7839 - accuracy: 0.8512 - val_loss: 0.7519 - val_accuracy: 0.8300\n",
            "Epoch 12/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.7212 - accuracy: 0.8600 - val_loss: 0.6934 - val_accuracy: 0.8450\n",
            "Epoch 13/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.6666 - accuracy: 0.8625 - val_loss: 0.6524 - val_accuracy: 0.8500\n",
            "Epoch 14/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.6202 - accuracy: 0.8675 - val_loss: 0.6170 - val_accuracy: 0.8600\n",
            "Epoch 15/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.5808 - accuracy: 0.8788 - val_loss: 0.5812 - val_accuracy: 0.8400\n",
            "Epoch 16/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.5469 - accuracy: 0.8775 - val_loss: 0.5525 - val_accuracy: 0.8450\n",
            "Epoch 17/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.5173 - accuracy: 0.8725 - val_loss: 0.5318 - val_accuracy: 0.8600\n",
            "Epoch 18/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.4920 - accuracy: 0.8788 - val_loss: 0.5034 - val_accuracy: 0.8500\n",
            "Epoch 19/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.4695 - accuracy: 0.8775 - val_loss: 0.4888 - val_accuracy: 0.8650\n",
            "Epoch 20/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.4495 - accuracy: 0.8825 - val_loss: 0.4732 - val_accuracy: 0.8600\n",
            "Epoch 21/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.4309 - accuracy: 0.8813 - val_loss: 0.4604 - val_accuracy: 0.8550\n",
            "Epoch 22/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.4144 - accuracy: 0.8875 - val_loss: 0.4421 - val_accuracy: 0.8650\n",
            "Epoch 23/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.3995 - accuracy: 0.8863 - val_loss: 0.4345 - val_accuracy: 0.8550\n",
            "Epoch 24/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.3863 - accuracy: 0.8900 - val_loss: 0.4259 - val_accuracy: 0.8700\n",
            "Epoch 25/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.3749 - accuracy: 0.8863 - val_loss: 0.4209 - val_accuracy: 0.8600\n",
            "Epoch 26/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.3631 - accuracy: 0.8963 - val_loss: 0.4070 - val_accuracy: 0.8700\n",
            "Epoch 27/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.3540 - accuracy: 0.8950 - val_loss: 0.4010 - val_accuracy: 0.8750\n",
            "Epoch 28/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.3427 - accuracy: 0.8950 - val_loss: 0.3943 - val_accuracy: 0.8700\n",
            "Epoch 29/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.3342 - accuracy: 0.9025 - val_loss: 0.3866 - val_accuracy: 0.8750\n",
            "Epoch 30/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.3255 - accuracy: 0.8963 - val_loss: 0.3812 - val_accuracy: 0.8850\n",
            "Epoch 31/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.3190 - accuracy: 0.8988 - val_loss: 0.3798 - val_accuracy: 0.8850\n",
            "Epoch 32/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.3110 - accuracy: 0.9050 - val_loss: 0.3741 - val_accuracy: 0.8900\n",
            "Epoch 33/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.3038 - accuracy: 0.9075 - val_loss: 0.3711 - val_accuracy: 0.8850\n",
            "Epoch 34/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.2983 - accuracy: 0.9112 - val_loss: 0.3625 - val_accuracy: 0.8850\n",
            "Epoch 35/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.2933 - accuracy: 0.9150 - val_loss: 0.3601 - val_accuracy: 0.8850\n",
            "Epoch 36/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.2857 - accuracy: 0.9137 - val_loss: 0.3559 - val_accuracy: 0.8900\n",
            "Epoch 37/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2812 - accuracy: 0.9175 - val_loss: 0.3554 - val_accuracy: 0.8900\n",
            "Epoch 38/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.2767 - accuracy: 0.9212 - val_loss: 0.3495 - val_accuracy: 0.8850\n",
            "Epoch 39/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2713 - accuracy: 0.9137 - val_loss: 0.3477 - val_accuracy: 0.8950\n",
            "Epoch 40/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.2666 - accuracy: 0.9200 - val_loss: 0.3519 - val_accuracy: 0.8850\n",
            "Epoch 41/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.2610 - accuracy: 0.9250 - val_loss: 0.3408 - val_accuracy: 0.8950\n",
            "Epoch 42/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.2583 - accuracy: 0.9250 - val_loss: 0.3412 - val_accuracy: 0.8900\n",
            "Epoch 43/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2532 - accuracy: 0.9212 - val_loss: 0.3399 - val_accuracy: 0.8950\n",
            "Epoch 44/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2490 - accuracy: 0.9212 - val_loss: 0.3347 - val_accuracy: 0.8850\n",
            "Epoch 45/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.2458 - accuracy: 0.9262 - val_loss: 0.3330 - val_accuracy: 0.8950\n",
            "Epoch 46/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2420 - accuracy: 0.9225 - val_loss: 0.3361 - val_accuracy: 0.8950\n",
            "Epoch 47/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.2382 - accuracy: 0.9300 - val_loss: 0.3310 - val_accuracy: 0.8950\n",
            "Epoch 48/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2347 - accuracy: 0.9287 - val_loss: 0.3244 - val_accuracy: 0.8950\n",
            "Epoch 49/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2313 - accuracy: 0.9312 - val_loss: 0.3268 - val_accuracy: 0.8950\n",
            "Epoch 50/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.2278 - accuracy: 0.9312 - val_loss: 0.3247 - val_accuracy: 0.8950\n",
            "Epoch 51/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2250 - accuracy: 0.9287 - val_loss: 0.3212 - val_accuracy: 0.8900\n",
            "Epoch 52/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.2226 - accuracy: 0.9275 - val_loss: 0.3186 - val_accuracy: 0.8950\n",
            "Epoch 53/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2194 - accuracy: 0.9287 - val_loss: 0.3174 - val_accuracy: 0.8950\n",
            "Epoch 54/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2166 - accuracy: 0.9300 - val_loss: 0.3169 - val_accuracy: 0.8950\n",
            "Epoch 55/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2128 - accuracy: 0.9287 - val_loss: 0.3152 - val_accuracy: 0.8950\n",
            "Epoch 56/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2112 - accuracy: 0.9275 - val_loss: 0.3154 - val_accuracy: 0.9000\n",
            "Epoch 57/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.2082 - accuracy: 0.9325 - val_loss: 0.3182 - val_accuracy: 0.8950\n",
            "Epoch 58/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2048 - accuracy: 0.9362 - val_loss: 0.3151 - val_accuracy: 0.8950\n",
            "Epoch 59/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.2025 - accuracy: 0.9350 - val_loss: 0.3097 - val_accuracy: 0.9000\n",
            "Epoch 60/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.2003 - accuracy: 0.9325 - val_loss: 0.3128 - val_accuracy: 0.8950\n",
            "Epoch 61/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1975 - accuracy: 0.9388 - val_loss: 0.3130 - val_accuracy: 0.8950\n",
            "Epoch 62/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.1962 - accuracy: 0.9350 - val_loss: 0.3106 - val_accuracy: 0.8850\n",
            "Epoch 63/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1938 - accuracy: 0.9337 - val_loss: 0.3089 - val_accuracy: 0.9000\n",
            "Epoch 64/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1907 - accuracy: 0.9413 - val_loss: 0.3069 - val_accuracy: 0.9000\n",
            "Epoch 65/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.1892 - accuracy: 0.9362 - val_loss: 0.3057 - val_accuracy: 0.9000\n",
            "Epoch 66/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.1868 - accuracy: 0.9425 - val_loss: 0.3056 - val_accuracy: 0.8950\n",
            "Epoch 67/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.1845 - accuracy: 0.9388 - val_loss: 0.3107 - val_accuracy: 0.8950\n",
            "Epoch 68/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1822 - accuracy: 0.9425 - val_loss: 0.3062 - val_accuracy: 0.8950\n",
            "Epoch 69/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1804 - accuracy: 0.9388 - val_loss: 0.3072 - val_accuracy: 0.8950\n",
            "Epoch 70/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1779 - accuracy: 0.9450 - val_loss: 0.3064 - val_accuracy: 0.8900\n",
            "Epoch 71/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1769 - accuracy: 0.9438 - val_loss: 0.3082 - val_accuracy: 0.8950\n",
            "Epoch 72/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1747 - accuracy: 0.9463 - val_loss: 0.3025 - val_accuracy: 0.8950\n",
            "Epoch 73/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.1720 - accuracy: 0.9438 - val_loss: 0.3065 - val_accuracy: 0.8900\n",
            "Epoch 74/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1709 - accuracy: 0.9475 - val_loss: 0.3046 - val_accuracy: 0.9000\n",
            "Epoch 75/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1683 - accuracy: 0.9513 - val_loss: 0.3082 - val_accuracy: 0.8900\n",
            "Epoch 76/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.1670 - accuracy: 0.9475 - val_loss: 0.3046 - val_accuracy: 0.8900\n",
            "Epoch 77/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.1654 - accuracy: 0.9488 - val_loss: 0.3069 - val_accuracy: 0.8950\n",
            "Epoch 78/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.1630 - accuracy: 0.9475 - val_loss: 0.3057 - val_accuracy: 0.8950\n",
            "Epoch 79/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1617 - accuracy: 0.9538 - val_loss: 0.3046 - val_accuracy: 0.8950\n",
            "Epoch 80/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1597 - accuracy: 0.9513 - val_loss: 0.3001 - val_accuracy: 0.8950\n",
            "Epoch 81/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1587 - accuracy: 0.9513 - val_loss: 0.3059 - val_accuracy: 0.8950\n",
            "Epoch 82/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1565 - accuracy: 0.9525 - val_loss: 0.3090 - val_accuracy: 0.8900\n",
            "Epoch 83/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.1554 - accuracy: 0.9588 - val_loss: 0.2994 - val_accuracy: 0.8950\n",
            "Epoch 84/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.1533 - accuracy: 0.9538 - val_loss: 0.3023 - val_accuracy: 0.8950\n",
            "Epoch 85/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1520 - accuracy: 0.9525 - val_loss: 0.3053 - val_accuracy: 0.8950\n",
            "Epoch 86/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1498 - accuracy: 0.9550 - val_loss: 0.3051 - val_accuracy: 0.8950\n",
            "Epoch 87/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.1501 - accuracy: 0.9588 - val_loss: 0.3005 - val_accuracy: 0.8950\n",
            "Epoch 88/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.1466 - accuracy: 0.9550 - val_loss: 0.3022 - val_accuracy: 0.8950\n",
            "Epoch 89/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1457 - accuracy: 0.9563 - val_loss: 0.3056 - val_accuracy: 0.8950\n",
            "Epoch 90/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.1431 - accuracy: 0.9625 - val_loss: 0.3031 - val_accuracy: 0.8950\n",
            "Epoch 91/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1434 - accuracy: 0.9600 - val_loss: 0.3036 - val_accuracy: 0.8950\n",
            "Epoch 92/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1408 - accuracy: 0.9588 - val_loss: 0.3040 - val_accuracy: 0.8950\n",
            "Epoch 93/100\n",
            "13/13 [==============================] - 0s 7ms/step - loss: 0.1402 - accuracy: 0.9625 - val_loss: 0.3026 - val_accuracy: 0.8950\n",
            "Epoch 94/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.1381 - accuracy: 0.9650 - val_loss: 0.3052 - val_accuracy: 0.8900\n",
            "Epoch 95/100\n",
            "13/13 [==============================] - 0s 6ms/step - loss: 0.1371 - accuracy: 0.9625 - val_loss: 0.3036 - val_accuracy: 0.8950\n",
            "Epoch 96/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.1356 - accuracy: 0.9638 - val_loss: 0.3076 - val_accuracy: 0.8900\n",
            "Epoch 97/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.1337 - accuracy: 0.9650 - val_loss: 0.3012 - val_accuracy: 0.8900\n",
            "Epoch 98/100\n",
            "13/13 [==============================] - 0s 5ms/step - loss: 0.1326 - accuracy: 0.9650 - val_loss: 0.3092 - val_accuracy: 0.9000\n",
            "Epoch 99/100\n",
            "13/13 [==============================] - 0s 8ms/step - loss: 0.1318 - accuracy: 0.9700 - val_loss: 0.3041 - val_accuracy: 0.8950\n",
            "Epoch 100/100\n",
            "13/13 [==============================] - 0s 9ms/step - loss: 0.1302 - accuracy: 0.9675 - val_loss: 0.3078 - val_accuracy: 0.8950\n"
          ]
        }
      ],
      "source": [
        "history = model.fit(x_train, y_train,\n",
        "          epochs=100,\n",
        "          batch_size=64, validation_split=0.2)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "history.history.keys()"
      ],
      "metadata": {
        "id": "PaLmJPBVqcnK",
        "outputId": "83b1ad43-614d-45df-da8d-2bc4b35d77c8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "id": "pSpOuK7Kn0UK"
      },
      "outputs": [],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['val_loss'])\n",
        "plt.plot(history.history['loss'])\n",
        "plt.legend(['val', 'train'])"
      ],
      "metadata": {
        "id": "S-aXXd_xq_tT",
        "outputId": "17755e66-7632-4557-f3cc-e802ea1e1fa1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 448
        }
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x797bd0454970>"
            ]
          },
          "metadata": {},
          "execution_count": 29
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "id": "t3QELHvWn0UL",
        "outputId": "ca314cb7-8461-44d1-d0af-d6102b12331d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 450
        }
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x797bde9580d0>"
            ]
          },
          "metadata": {},
          "execution_count": 28
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "plt.plot(history.history['val_accuracy'])\n",
        "plt.plot(history.history['accuracy'])\n",
        "plt.legend(['val', 'train'])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-b_6qvBFn0UL"
      },
      "source": [
        "## <div style=\"direction:rtl;text-align:right;font-family:B Lotus, B Nazanin, Tahoma\">ارزیابی مدل روی داده های آزمون</div>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "id": "zNl9cUA2n0UL",
        "outputId": "727bcec2-33aa-4e04-88b8-800799a448ed",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "7/7 [==============================] - 0s 3ms/step - loss: 0.2483 - accuracy: 0.9250\n",
            "\n",
            "Testing loss: 0.25, acc: 0.93%\n"
          ]
        }
      ],
      "source": [
        "loss, acc = model.evaluate(x_test, y_test)\n",
        "print('\\nTesting loss: %.2f, acc: %.2f%%'%(loss, acc))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ruUxAbvWn0UM"
      },
      "source": [
        "## <div style=\"direction:rtl;text-align:right;font-family:B Lotus, B Nazanin, Tahoma\">پیش‌بینی داده‌های آموزشی</div>\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# The predict_classes function outputs the highest probability class\n",
        "# according to the trained classifier for each input example.\n",
        "predicted = model.predict(x_test)"
      ],
      "metadata": {
        "id": "JVzqr5A-rNYp",
        "outputId": "2673ebb9-57c6-4410-ffe7-c04b429ba292",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "7/7 [==============================] - 0s 3ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "predicted_classes = np.argmax(predicted, axis=1)"
      ],
      "metadata": {
        "id": "G2lTnueerrU0"
      },
      "execution_count": 39,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": 40,
      "metadata": {
        "id": "ZlYHuZvkn0UO",
        "outputId": "4c9335bf-2212-4a95-8946-1a97f4f40afd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "predicted:\n",
            "[7 2 3 1 5 5 4 7 3 2 0 8 8 0 2 9 3 6 7 4 0 3 6 3 9 2 7 5 2 9 2 5 5 8 9 6 5\n",
            " 1 4 8 8 4 7 2 1 2 7 9 0 3 7 5 7 5 2 9 8 2 9 8 8 6 6 6 6 6 2 4 2 4 4 5 9 1\n",
            " 8 4 0 5 6 2 4 3 2 7 7 7 7 1 8 1 7 8 7 7 8 9 7 2 3 1 0 2 9 6 3 5 5 0 0 9 6\n",
            " 7 9 3 9 9 8 7 9 2 5 2 5 5 9 6 9 2 5 3 7 9 5 2 9 0 4 1 8 2 2 3 5 2 9 3 8 2\n",
            " 7 0 9 9 0 7 6 5 4 4 9 3 7 0 7 4 9 4 7 3 4 1 5 6 7 9 1 3 5 4 5 3 4 1 3 3 1\n",
            " 1 3 3 8 9 6 7 7 2 3 0 1 4 9 5]\n",
            "True Label:\n",
            "[7 2 3 1 5 5 4 7 3 2 0 8 8 0 2 9 3 6 7 4 0 3 6 3 9 2 7 5 2 9 7 5 5 8 9 6 5\n",
            " 1 4 8 8 4 7 7 1 2 7 9 0 3 7 4 7 5 2 9 8 2 9 8 8 6 6 6 6 6 2 4 3 4 4 5 9 1\n",
            " 8 2 0 5 6 2 4 3 2 7 7 7 7 1 8 1 7 8 7 7 8 9 3 2 3 1 0 2 9 6 3 5 5 0 0 3 6\n",
            " 7 9 3 9 9 8 7 9 2 5 2 5 5 9 6 9 2 0 3 7 6 5 2 9 0 4 1 8 2 2 3 0 2 9 3 8 6\n",
            " 7 0 9 9 0 7 6 5 4 7 9 3 7 0 7 1 9 4 7 3 4 1 5 6 7 9 1 3 5 4 5 7 4 1 3 3 1\n",
            " 2 3 3 8 9 6 7 7 2 3 0 1 4 9 5]\n"
          ]
        }
      ],
      "source": [
        "\n",
        "print(\"predicted:\")\n",
        "print(predicted_classes)\n",
        "print(\"True Label:\")\n",
        "print(y_test_original)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 41,
      "metadata": {
        "id": "8mieKbW2n0UP",
        "outputId": "d370b4f0-e114-46b8-907c-fa3acdd33237",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 457
        }
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 480x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "plt.matshow(confusion_matrix(y_test_original, predicted_classes), cmap=\"Blues\")\n",
        "plt.xticks(range(10))\n",
        "plt.yticks(range(10))\n",
        "plt.xlabel(\"Predicted label\")\n",
        "plt.ylabel(\"True label\");"
      ]
    }
  ],
  "metadata": {
    "anaconda-cloud": {},
    "kernelspec": {
      "display_name": "tf2.5",
      "language": "python",
      "name": "tensorflow_25"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.10"
    },
    "nbpresent": {
      "slides": {
        "300ee14f-a043-486e-b274-7ff253907cd7": {
          "id": "300ee14f-a043-486e-b274-7ff253907cd7",
          "prev": "cb74e0bc-4513-4d13-b7f1-14c3078a7927",
          "regions": {
            "26dc3f39-a230-447c-af4c-f5e5b2fb7835": {
              "attrs": {
                "height": 0.8,
                "width": 0.8,
                "x": 0.1,
                "y": 0.1
              },
              "content": {
                "cell": "c58440a5-3f8f-4f37-9c79-6bf766209406",
                "part": "whole"
              },
              "id": "26dc3f39-a230-447c-af4c-f5e5b2fb7835"
            }
          }
        },
        "878aa53a-1444-4100-8f50-7a408191c579": {
          "id": "878aa53a-1444-4100-8f50-7a408191c579",
          "prev": null,
          "regions": {
            "a6c6843a-5ea6-4fbc-b890-3b4b8ae475b3": {
              "attrs": {
                "height": 0.8,
                "width": 0.8,
                "x": 0.1,
                "y": 0.1
              },
              "content": {
                "cell": "588ee1fa-64b5-453b-ade7-8e6b2515821c",
                "part": "whole"
              },
              "id": "a6c6843a-5ea6-4fbc-b890-3b4b8ae475b3"
            }
          }
        },
        "96ffe88e-7b50-43de-afdd-942e564f4e3e": {
          "id": "96ffe88e-7b50-43de-afdd-942e564f4e3e",
          "prev": "878aa53a-1444-4100-8f50-7a408191c579",
          "regions": {
            "b7e52e12-489a-468d-b10c-af2024fd2856": {
              "attrs": {
                "height": 0.8,
                "width": 0.8,
                "x": 0.1,
                "y": 0.1
              },
              "content": {
                "cell": "de829a92-1fb6-44ad-a2c6-fc1001e1f6e1",
                "part": "whole"
              },
              "id": "b7e52e12-489a-468d-b10c-af2024fd2856"
            }
          }
        },
        "cb74e0bc-4513-4d13-b7f1-14c3078a7927": {
          "id": "cb74e0bc-4513-4d13-b7f1-14c3078a7927",
          "prev": "96ffe88e-7b50-43de-afdd-942e564f4e3e",
          "regions": {
            "444878ee-68f3-4abb-acff-a7079b21e86d": {
              "attrs": {
                "height": 0.8,
                "width": 0.8,
                "x": 0.1,
                "y": 0.1
              },
              "content": {
                "cell": "25f3f538-1ee8-4d98-a6bb-14cbeb7a702d",
                "part": "whole"
              },
              "id": "444878ee-68f3-4abb-acff-a7079b21e86d"
            }
          }
        }
      },
      "themes": {}
    },
    "colab": {
      "provenance": [],
      "gpuType": "T4"
    },
    "accelerator": "GPU"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}