nemolize/minist-10-samples.ipynb

## minist-10-samples.ipynb
{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyO/179XH0EiL9kqKvdWhjUG",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/nemolize/c4f7e1439791a99beeb95673f841a629/minist-10-samples.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "id": "sF30swAUYBYA"
      },
      "outputs": [],
      "source": [
        "# TensorFlowとkerasのインポート\n",
        "from tensorflow import keras\n",
        "\n",
        "# MINISTデータの読み込み\n",
        "(x_train, y_train), (x_test, y_test) = keras.datasets.mnist.load_data()\n",
        "\n",
        "# データの正規化\n",
        "x_train = x_train / 255.0\n",
        "x_test = x_test / 255.0\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# モデルの定義\n",
        "model = keras.Sequential([\n",
        "  keras.layers.Flatten(input_shape=(28, 28)),\n",
        "  keras.layers.Dense(128, activation='relu'),\n",
        "  keras.layers.Dense(10, activation='softmax')\n",
        "])\n",
        "\n",
        "# モデルのコンパイル\n",
        "model.compile(\n",
        "  optimizer='adam',\n",
        "  loss='sparse_categorical_crossentropy',\n",
        "  metrics=['accuracy']\n",
        ")\n",
        "\n",
        "# モデルの訓練\n",
        "model.fit(x_train, y_train, epochs=5)"
      ],
      "metadata": {
        "id": "UBEsTvadaihV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# テストデータから10個のサンプルを選択\n",
        "num_samples = 10\n",
        "indices = np.random.choice(len(x_test), num_samples)\n",
        "\n",
        "# 予測を実行\n",
        "predictions = model.predict(x_test[indices])\n",
        "\n",
        "# 結果を表示\n",
        "for i, index in enumerate(indices):\n",
        "    plt.subplot(2, 5, i + 1)\n",
        "    plt.imshow(x_test[index], cmap='gray')\n",
        "    plt.axis('off')\n",
        "    plt.title(f'Predicted: {np.argmax(predictions[i])}\\nTrue: {y_test[index]}')\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 420
        },
        "id": "PoL01RjdY_XP",
        "outputId": "f8779344-a85d-4d41-f546-25dfbc2e1098"
      },
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 45ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 10 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}