ANN_cifar10(week9).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "mount_file_id": "15b2ec4jqbQ2I1luwoQzdsJ2eBdC36rXP",
      "authorship_tag": "ABX9TyO5Rc6RcNYV8iexEbqLnOQi",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "gpuClass": "standard",
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/thanhtruongphan/Artificial-Intelligence-course/blob/main/ANN_cifar10(week9).ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## **Quy trình làm AI mạng ANN**"
      ],
      "metadata": {
        "id": "NHU_SkJHzhV4"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "*   data\n",
        "*   model\n",
        "*   compile\n",
        "*   train\n",
        "*   evaluate\n",
        "\n"
      ],
      "metadata": {
        "id": "wxXomJIiZErz"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "D3adwEAwhU4n"
      },
      "outputs": [],
      "source": [
        "from keras.datasets import cifar10\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# keras.datasets.cifar10.load_data"
      ],
      "metadata": {
        "id": "tA9ZlivBhloh"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "(X_train, y_train),(X_test,y_test) = cifar10.load_data()"
      ],
      "metadata": {
        "id": "oRdcMm29iLqf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_train.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DyYsDcasiT8X",
        "outputId": "c98bd3cc-7cc2-44bd-96c5-f2735d805ddf"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(50000, 32, 32, 3)"
            ]
          },
          "metadata": {},
          "execution_count": 33
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_train.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wy9ieEkFiUuh",
        "outputId": "b6ed0b54-de3d-4974-bd42-0d0f6ae5c18c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(50000, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 34
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_test.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2EenU2CBizdI",
        "outputId": "1afdd644-0593-45d5-a26f-29dfdd68b17d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10000, 32, 32, 3)"
            ]
          },
          "metadata": {},
          "execution_count": 35
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_test.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EhnoXzTBi2Cr",
        "outputId": "1de24b95-3f01-4384-ab0a-355f76c3c14a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10000, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 36
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "digit = X_train[15]\n",
        "plt.imshow(digit, cmap=plt.cm.binary)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "L54fERlyi3uB",
        "outputId": "36b22315-88c1-48ae-8b47-602d26250fde"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print('Mo hinh so : ',y_train[15])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TBL2JcBfi9hT",
        "outputId": "2cdc5aaf-b22e-4556-a871-8ff8e3d034df"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mo hinh so :  [9]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for i in range(15):\n",
        "  plt.subplot(3,5,i+1)\n",
        "  plt.tight_layout()\n",
        "  plt.imshow(X_train[i],cmap='gray',interpolation='none')\n",
        "  plt.title('Mo hinh so:{}'.format(y_train[i]))\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 481
        },
        "id": "ju3SHV7CjHE4",
        "outputId": "d508f623-1f90-415b-a220-50e8c7eddba3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-39-93b9fca45055>:2: MatplotlibDeprecationWarning: Auto-removal of overlapping axes is deprecated since 3.6 and will be removed two minor releases later; explicitly call ax.remove() as needed.\n",
            "  plt.subplot(3,5,i+1)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 14 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "t-1inG6EjVgG"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## **ve mang**"
      ],
      "metadata": {
        "id": "ZRj4iInyjnzF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "(X_train, y_train),(X_test,y_test) = cifar10.load_data()"
      ],
      "metadata": {
        "id": "FQdCH8lj5RmH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from keras.utils import to_categorical"
      ],
      "metadata": {
        "id": "xQdCCuRImgP1"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from keras.models import Sequential \n",
        "from keras.layers import Dense \n",
        "model = Sequential()\n",
        "# model.add(Dense(512,activation='relu',input_shape=(32*32*3,)))\n",
        "model.add(Dense(5000,activation='relu',input_shape=(32*32*3,)))\n",
        "model.add(Dense(3000,activation='relu'))\n",
        "model.add(Dense(1000,activation='relu'))\n",
        "model.add(Dense(10,activation='softmax'))\n",
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ohSGIKt1jqiY",
        "outputId": "dc290fd6-3e9e-4524-b9da-e95671bf40d7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_9\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_36 (Dense)            (None, 5000)              15365000  \n",
            "                                                                 \n",
            " dense_37 (Dense)            (None, 3000)              15003000  \n",
            "                                                                 \n",
            " dense_38 (Dense)            (None, 1000)              3001000   \n",
            "                                                                 \n",
            " dense_39 (Dense)            (None, 10)                10010     \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 33,379,010\n",
            "Trainable params: 33,379,010\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X = X_train\n",
        "y = y_train\n",
        "# \n",
        "X_train = X_train.reshape(50000,32*32*3)\n",
        "X_train = X_train.astype('float32')/255\n",
        "X_test = X_test.reshape(10000,32*32*3)\n",
        "X_test = X_test.astype('float32')/255\n",
        "#\n",
        "X_train.shape"
      ],
      "metadata": {
        "id": "PbuR2kFbkKAm",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6a90441c-fa41-4575-9ed9-b62bcafaa171"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(50000, 3072)"
            ]
          },
          "metadata": {},
          "execution_count": 79
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_train = to_categorical(y_train,10)  #1 to 10 bits\n",
        "y_test = to_categorical(y_test,10)\n",
        "y_train.shape\n",
        "y_test.shape"
      ],
      "metadata": {
        "id": "kxQK1rEsYLgd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ebde9267-f296-4a53-98a8-9fdff19243e4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10000, 10)"
            ]
          },
          "metadata": {},
          "execution_count": 80
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(optimizer='rmsprop',loss='categorical_crossentropy',metrics=['accuracy'])\n",
        "model.fit(X_train,y_train,epochs=100,batch_size=128)"
      ],
      "metadata": {
        "id": "pWNFwgL2ZSCc",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1be03c45-d680-4ea2-db24-86acad662e78"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "391/391 [==============================] - 6s 11ms/step - loss: 2.4265 - accuracy: 0.2469\n",
            "Epoch 2/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.8171 - accuracy: 0.3464\n",
            "Epoch 3/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.7242 - accuracy: 0.3821\n",
            "Epoch 4/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.6537 - accuracy: 0.4094\n",
            "Epoch 5/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.5923 - accuracy: 0.4320\n",
            "Epoch 6/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.5524 - accuracy: 0.4439\n",
            "Epoch 7/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.5096 - accuracy: 0.4630\n",
            "Epoch 8/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.4777 - accuracy: 0.4751\n",
            "Epoch 9/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.4454 - accuracy: 0.4835\n",
            "Epoch 10/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.4108 - accuracy: 0.4967\n",
            "Epoch 11/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.3830 - accuracy: 0.5061\n",
            "Epoch 12/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.3492 - accuracy: 0.5194\n",
            "Epoch 13/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.3208 - accuracy: 0.5315\n",
            "Epoch 14/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.2886 - accuracy: 0.5431\n",
            "Epoch 15/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.2470 - accuracy: 0.5567\n",
            "Epoch 16/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.2247 - accuracy: 0.5641\n",
            "Epoch 17/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.1887 - accuracy: 0.5795\n",
            "Epoch 18/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.1560 - accuracy: 0.5903\n",
            "Epoch 19/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.1215 - accuracy: 0.6016\n",
            "Epoch 20/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.0868 - accuracy: 0.6153\n",
            "Epoch 21/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.0558 - accuracy: 0.6300\n",
            "Epoch 22/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 1.0247 - accuracy: 0.6422\n",
            "Epoch 23/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.9911 - accuracy: 0.6559\n",
            "Epoch 24/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.9583 - accuracy: 0.6671\n",
            "Epoch 25/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.9281 - accuracy: 0.6791\n",
            "Epoch 26/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.8928 - accuracy: 0.6912\n",
            "Epoch 27/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.8744 - accuracy: 0.6993\n",
            "Epoch 28/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.8379 - accuracy: 0.7119\n",
            "Epoch 29/100\n",
            "391/391 [==============================] - 5s 12ms/step - loss: 0.8141 - accuracy: 0.7219\n",
            "Epoch 30/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.7878 - accuracy: 0.7309\n",
            "Epoch 31/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.7651 - accuracy: 0.7424\n",
            "Epoch 32/100\n",
            "391/391 [==============================] - 5s 12ms/step - loss: 0.7399 - accuracy: 0.7518\n",
            "Epoch 33/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.7134 - accuracy: 0.7619\n",
            "Epoch 34/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.6841 - accuracy: 0.7743\n",
            "Epoch 35/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.6655 - accuracy: 0.7782\n",
            "Epoch 36/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.6574 - accuracy: 0.7827\n",
            "Epoch 37/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.6364 - accuracy: 0.7933\n",
            "Epoch 38/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.6204 - accuracy: 0.7976\n",
            "Epoch 39/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.6176 - accuracy: 0.8016\n",
            "Epoch 40/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.6017 - accuracy: 0.8083\n",
            "Epoch 41/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.5803 - accuracy: 0.8167\n",
            "Epoch 42/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.5680 - accuracy: 0.8186\n",
            "Epoch 43/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.5654 - accuracy: 0.8247\n",
            "Epoch 44/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.5463 - accuracy: 0.8310\n",
            "Epoch 45/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.5381 - accuracy: 0.8343\n",
            "Epoch 46/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.5368 - accuracy: 0.8377\n",
            "Epoch 47/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.5159 - accuracy: 0.8415\n",
            "Epoch 48/100\n",
            "391/391 [==============================] - 5s 12ms/step - loss: 0.5169 - accuracy: 0.8450\n",
            "Epoch 49/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.5070 - accuracy: 0.8500\n",
            "Epoch 50/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.5196 - accuracy: 0.8455\n",
            "Epoch 51/100\n",
            "391/391 [==============================] - 5s 12ms/step - loss: 0.4998 - accuracy: 0.8544\n",
            "Epoch 52/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4833 - accuracy: 0.8579\n",
            "Epoch 53/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4708 - accuracy: 0.8611\n",
            "Epoch 54/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4840 - accuracy: 0.8606\n",
            "Epoch 55/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4920 - accuracy: 0.8615\n",
            "Epoch 56/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4643 - accuracy: 0.8676\n",
            "Epoch 57/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4662 - accuracy: 0.8703\n",
            "Epoch 58/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4514 - accuracy: 0.8748\n",
            "Epoch 59/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4479 - accuracy: 0.8752\n",
            "Epoch 60/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4490 - accuracy: 0.8768\n",
            "Epoch 61/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4436 - accuracy: 0.8793\n",
            "Epoch 62/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4469 - accuracy: 0.8794\n",
            "Epoch 63/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4457 - accuracy: 0.8817\n",
            "Epoch 64/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4315 - accuracy: 0.8851\n",
            "Epoch 65/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4228 - accuracy: 0.8867\n",
            "Epoch 66/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4192 - accuracy: 0.8885\n",
            "Epoch 67/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4242 - accuracy: 0.8886\n",
            "Epoch 68/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4070 - accuracy: 0.8941\n",
            "Epoch 69/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4146 - accuracy: 0.8929\n",
            "Epoch 70/100\n",
            "391/391 [==============================] - 5s 12ms/step - loss: 0.4138 - accuracy: 0.8954\n",
            "Epoch 71/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4011 - accuracy: 0.8973\n",
            "Epoch 72/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3961 - accuracy: 0.8985\n",
            "Epoch 73/100\n",
            "391/391 [==============================] - 5s 12ms/step - loss: 0.4070 - accuracy: 0.8976\n",
            "Epoch 74/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3931 - accuracy: 0.9053\n",
            "Epoch 75/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3895 - accuracy: 0.9034\n",
            "Epoch 76/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3808 - accuracy: 0.9054\n",
            "Epoch 77/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.4018 - accuracy: 0.9046\n",
            "Epoch 78/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3965 - accuracy: 0.9062\n",
            "Epoch 79/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3944 - accuracy: 0.9068\n",
            "Epoch 80/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3776 - accuracy: 0.9117\n",
            "Epoch 81/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3813 - accuracy: 0.9086\n",
            "Epoch 82/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3883 - accuracy: 0.9112\n",
            "Epoch 83/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3681 - accuracy: 0.9147\n",
            "Epoch 84/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3789 - accuracy: 0.9140\n",
            "Epoch 85/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3770 - accuracy: 0.9144\n",
            "Epoch 86/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3805 - accuracy: 0.9154\n",
            "Epoch 87/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3659 - accuracy: 0.9182\n",
            "Epoch 88/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3747 - accuracy: 0.9183\n",
            "Epoch 89/100\n",
            "391/391 [==============================] - 5s 12ms/step - loss: 0.3605 - accuracy: 0.9203\n",
            "Epoch 90/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3743 - accuracy: 0.9187\n",
            "Epoch 91/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3641 - accuracy: 0.9206\n",
            "Epoch 92/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3623 - accuracy: 0.9212\n",
            "Epoch 93/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3373 - accuracy: 0.9256\n",
            "Epoch 94/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3615 - accuracy: 0.9229\n",
            "Epoch 95/100\n",
            "391/391 [==============================] - 5s 12ms/step - loss: 0.3613 - accuracy: 0.9247\n",
            "Epoch 96/100\n",
            "391/391 [==============================] - 5s 12ms/step - loss: 0.3610 - accuracy: 0.9254\n",
            "Epoch 97/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3545 - accuracy: 0.9259\n",
            "Epoch 98/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3502 - accuracy: 0.9281\n",
            "Epoch 99/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3453 - accuracy: 0.9279\n",
            "Epoch 100/100\n",
            "391/391 [==============================] - 4s 11ms/step - loss: 0.3574 - accuracy: 0.9276\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f0fe5a1d7c0>"
            ]
          },
          "metadata": {},
          "execution_count": 81
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_loss,test_acc = model.evaluate(X_test,y_test)\n",
        "model.save('/content/drive/MyDrive/A_HOC TAP/Nam 4, Hoc ky 2/Artificial Intelligence/Code_AI/ipython/mang/model_ANN_cifar10.h5')\n",
        "print('Mat mat: ',test_loss)\n",
        "print('Do chinh xac: ', test_acc)"
      ],
      "metadata": {
        "id": "THNoLHzPZxfk",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4d051aa4-3deb-4cb6-f672-df0c74fb5222"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 1s 3ms/step - loss: 11.0708 - accuracy: 0.4872\n",
            "Mat mat:  11.070837020874023\n",
            "Do chinh xac:  0.487199991941452\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "tdN8OwPwbsx5"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "digit = X[100]\n",
        "plt.imshow(digit, cmap=plt.cm.binary)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "5bbHXtmzclV6",
        "outputId": "ba0c7735-54d6-4deb-a9ab-5540960270d2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_pred = X[555].reshape(1,32*32*3)\n",
        "X_pred = X_pred.astype('float32')/255\n",
        "y_pred = model.predict(X_pred)\n",
        "print(y_pred)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zvOxA8Fyc1-Q",
        "outputId": "64889240-1fa4-4f9e-9aa8-12431d97271c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 87ms/step\n",
            "[[1.0000000e+00 5.0985571e-34 4.0904077e-32 1.9585061e-19 0.0000000e+00\n",
            "  1.0754190e-33 0.0000000e+00 0.0000000e+00 1.9221941e-26 2.8182648e-15]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# 9.9981e-01 lon nhat - dem thay o vi tri so 4"
      ],
      "metadata": {
        "id": "8aZXRmd5fmNR"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# digit = X[555]\n",
        "# plt.imshow(digit, cmap=plt.cm.binary)\n",
        "# plt.show()"
      ],
      "metadata": {
        "id": "Y2WYKz8SgLJ6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# np.argmax(model.predict(X_pred),axis=1)"
      ],
      "metadata": {
        "id": "5PWtepGzgPSP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "MPj4H72spOMO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**model.save('test1.h5')**"
      ],
      "metadata": {
        "id": "q-5HwaMJpYM-"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**open**"
      ],
      "metadata": {
        "id": "YT7zUTVwpfLt"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**model10 = load('test1.h5')**"
      ],
      "metadata": {
        "id": "GquTgwmEpi-W"
      }
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "DY_oJjXeqNIw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "Vk9ER8z3qNWX"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from keras.utils import load_img\n",
        "img = load_img('/content/drive/MyDrive/A_HOC TAP/Nam 4, Hoc ky 2/Artificial Intelligence/Code_AI/files/images/meo4.jpg',target_size=(32,32))\n",
        "plt.imshow(img)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 448
        },
        "id": "aqTOb4d0qNdH",
        "outputId": "b33a6b1f-2219-435a-adbe-bdb530438b5c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f0ff15467f0>"
            ]
          },
          "metadata": {},
          "execution_count": 85
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow.keras.utils import img_to_array\n",
        "\n",
        "\n",
        "img = img_to_array(img)\n",
        "img = img.reshape(1,32*32*3)\n",
        "img = img.astype('float32')\n",
        "img = img/255\n",
        "np.argmax(model.predict(img),axis=-1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7ULvbEglq77W",
        "outputId": "f7b47ddb-3974-4012-f0b0-ebb928bf80b5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 19ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([3])"
            ]
          },
          "metadata": {},
          "execution_count": 86
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from keras.utils import load_img\n",
        "img = load_img('/content/drive/MyDrive/A_HOC TAP/Nam 4, Hoc ky 2/Artificial Intelligence/Code_AI/files/images/meo2.jpg',target_size=(32,32))\n",
        "plt.imshow(img)"
      ],
      "metadata": {
        "id": "CPPssx9srUZj",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 448
        },
        "outputId": "edb96ad4-e5e1-4c4c-8d9b-bdacdb9fc320"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f0fec5c8ee0>"
            ]
          },
          "metadata": {},
          "execution_count": 89
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow.keras.utils import img_to_array\n",
        "\n",
        "\n",
        "img = img_to_array(img)\n",
        "img = img.reshape(1,32*32*3)\n",
        "img = img.astype('float32')\n",
        "img = img/255\n",
        "np.argmax(model.predict(img),axis=-1)"
      ],
      "metadata": {
        "id": "dd9_fNnZr_VJ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d8610f40-2f4e-4d58-bcc5-0673ddc94774"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 17ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([2])"
            ]
          },
          "metadata": {},
          "execution_count": 90
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "dq2JV8aW_tzs"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}