machine_learning_13_1.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "toc_visible": true,
      "authorship_tag": "ABX9TyPj5CCnXQUsh+QOAaqGZlTJ",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/matsunagalab/lecture_ML/blob/main/machine_learning_13_1.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 第13回 ニューラルネットワークそのe: 代表的なニューラルネット CNN"
      ],
      "metadata": {
        "id": "Y3pXof4n2o2j"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## MNIST dataset (digits dataset)をConvolutional Neural Network (CNN)で学習"
      ],
      "metadata": {
        "id": "vjo4NNqaVeIj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Load requied packages\n",
        "import torch\n",
        "import torchvision\n",
        "from sklearn import datasets\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "# Load and preprocess the data\n",
        "digits = datasets.load_digits()\n",
        "data = digits.images\n",
        "labels = digits.target\n",
        "\n",
        "# Preprocess the data\n",
        "scaler = MinMaxScaler()\n",
        "data = scaler.fit_transform(data.reshape(-1, 64)).reshape(-1, 8, 8)\n",
        "\n",
        "# Split the data into train and test datasets\n",
        "train_data, test_data, train_labels, test_labels = train_test_split(data, labels, test_size=0.2, random_state=42)\n",
        "\n",
        "# Convert the data to PyTorch tensors\n",
        "train_data = torch.tensor(train_data, dtype=torch.float32).unsqueeze(1)\n",
        "test_data = torch.tensor(test_data, dtype=torch.float32).unsqueeze(1)\n",
        "train_labels = torch.tensor(train_labels, dtype=torch.long)\n",
        "test_labels = torch.tensor(test_labels, dtype=torch.long)\n",
        "\n",
        "# Define the CNN\n",
        "model = torch.nn.Sequential(\n",
        "    torch.nn.Conv2d(1, 32, kernel_size=3, stride=1, padding=1),\n",
        "    torch.nn.ReLU(),\n",
        "    torch.nn.MaxPool2d(kernel_size=2, stride=2),\n",
        "    torch.nn.Conv2d(32, 64, kernel_size=3, stride=1, padding=1),\n",
        "    torch.nn.ReLU(),\n",
        "    torch.nn.MaxPool2d(kernel_size=2, stride=2),\n",
        "    torch.nn.Flatten(),\n",
        "    torch.nn.Linear(64*2*2, 10),\n",
        ")\n",
        "\n",
        "# Move the model to GPU if available\n",
        "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
        "model = model.to(device)\n",
        "\n",
        "# Define the loss function and the optimizer\n",
        "loss_fn = torch.nn.CrossEntropyLoss()\n",
        "optimizer = torch.optim.Adam(model.parameters())\n",
        "\n",
        "# Train the model\n",
        "losses = []\n",
        "epochs = 200\n",
        "for epoch in range(epochs):\n",
        "    inputs = train_data.to(device)\n",
        "    labels = train_labels.to(device)\n",
        "    optimizer.zero_grad()\n",
        "    outputs = model(inputs)\n",
        "    loss = loss_fn(outputs, labels)\n",
        "    loss.backward()\n",
        "    optimizer.step()\n",
        "\n",
        "    # Record the loss\n",
        "    losses.append(loss.item())\n",
        "\n",
        "    if epoch % 10 == 0:\n",
        "        print(f\"Epoch: {epoch}, Loss: {loss.item()}\")\n",
        "\n",
        "# Test the model\n",
        "model.eval()\n",
        "test_data = test_data.to(device)\n",
        "test_labels = test_labels.to(device)\n",
        "outputs = model(test_data)\n",
        "_, predicted = torch.max(outputs.data, 1)\n",
        "correct = (predicted == test_labels).sum().item()\n",
        "print('Accuracy: {:.2f} %'.format(correct / len(test_labels) * 100))\n",
        "\n",
        "# Plot the loss during training\n",
        "plt.figure(figsize=(9,5))\n",
        "plt.plot(losses)\n",
        "plt.xlabel(\"Epoch\")\n",
        "plt.ylabel(\"Loss\")\n",
        "plt.show()\n",
        "\n",
        "# Plot the results\n",
        "fig, axs = plt.subplots(2, 5, figsize=(10, 5))\n",
        "for i in range(10):\n",
        "    ax = axs[i//5, i%5]\n",
        "    ax.imshow(test_data[i, 0].cpu().numpy(), cmap='gray')\n",
        "    ax.set_title(f\"True:{test_labels[i]}, Pred:{predicted[i]}\")\n",
        "plt.show()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "LcJYicl8iyKV",
        "outputId": "1fd5fdf8-0451-4115-9f00-42db9c9c046d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 0, Loss: 2.295428991317749\n",
            "Epoch: 10, Loss: 2.098356008529663\n",
            "Epoch: 20, Loss: 1.7881120443344116\n",
            "Epoch: 30, Loss: 1.3279504776000977\n",
            "Epoch: 40, Loss: 0.8380991816520691\n",
            "Epoch: 50, Loss: 0.5011556148529053\n",
            "Epoch: 60, Loss: 0.32201412320137024\n",
            "Epoch: 70, Loss: 0.22579821944236755\n",
            "Epoch: 80, Loss: 0.16916602849960327\n",
            "Epoch: 90, Loss: 0.13249430060386658\n",
            "Epoch: 100, Loss: 0.10683083534240723\n",
            "Epoch: 110, Loss: 0.08771676570177078\n",
            "Epoch: 120, Loss: 0.07288511097431183\n",
            "Epoch: 130, Loss: 0.061176929622888565\n",
            "Epoch: 140, Loss: 0.05174688622355461\n",
            "Epoch: 150, Loss: 0.04408714547753334\n",
            "Epoch: 160, Loss: 0.0377906896173954\n",
            "Epoch: 170, Loss: 0.03261176496744156\n",
            "Epoch: 180, Loss: 0.02832828462123871\n",
            "Epoch: 190, Loss: 0.02475799061357975\n",
            "Accuracy: 99.17 %\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 900x500 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x500 with 10 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "Lea1kxvfOuDT"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}