{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# 多分类问题 - 手写数字识别\n", "\n", "## 数据集\n", "- minst数据集(手写数字数据集)\n", "\n", "## 激活函数\n", "- softmax\n", "\n", "## 损失函数\n", "- 交叉熵\n", "\n", "## 优化器\n", "- 梯度下降\n", "\n", "## 模型\n", "- 全连接层\n" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [], "source": [ "# 导库\n", "import tensorflow as tf\n", "from tensorflow.keras import Sequential\n", "from tensorflow.keras.layers import Dense\n", "from tensorflow.keras.losses import SparseCategoricalCrossentropy" ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
Model: \"sequential_2\"\n",
       "
\n" ], "text/plain": [ "\u001b[1mModel: \"sequential_2\"\u001b[0m\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃ Layer (type)                    ┃ Output Shape           ┃       Param # ┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ dense_4 (Dense)                 │ ?                      │   0 (unbuilt) │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_5 (Dense)                 │ ?                      │   0 (unbuilt) │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_6 (Dense)                 │ ?                      │   0 (unbuilt) │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
       "
\n" ], "text/plain": [ "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n", "┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n", "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n", "│ dense_4 (\u001b[38;5;33mDense\u001b[0m) │ ? │ \u001b[38;5;34m0\u001b[0m (unbuilt) │\n", "├─────────────────────────────────┼────────────────────────┼───────────────┤\n", "│ dense_5 (\u001b[38;5;33mDense\u001b[0m) │ ? │ \u001b[38;5;34m0\u001b[0m (unbuilt) │\n", "├─────────────────────────────────┼────────────────────────┼───────────────┤\n", "│ dense_6 (\u001b[38;5;33mDense\u001b[0m) │ ? │ \u001b[38;5;34m0\u001b[0m (unbuilt) │\n", "└─────────────────────────────────┴────────────────────────┴───────────────┘\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Total params: 0 (0.00 B)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Trainable params: 0 (0.00 B)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Non-trainable params: 0 (0.00 B)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "None\n" ] } ], "source": [ "# TEST\n", "model = Sequential([Dense(units=25,activation='relu'),\n", " Dense(units=15,activation='relu'),\n", " Dense(units=10,activation='softmax')])\n", "model.compile(loss=SparseCategoricalCrossentropy())\n", "# 输出模型\n", "print(model.summary())" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/5\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "2025-01-21 11:03:07.974903: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:117] Plugin optimizer for device_type GPU is enabled.\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "\u001b[1m1875/1875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m16s\u001b[0m 8ms/step - accuracy: 0.8340 - loss: 0.5514\n", "Epoch 2/5\n", "\u001b[1m1875/1875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m16s\u001b[0m 8ms/step - accuracy: 0.8933 - loss: 0.3836\n", "Epoch 3/5\n", "\u001b[1m1875/1875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 8ms/step - accuracy: 0.8902 - loss: 0.4045\n", "Epoch 4/5\n", "\u001b[1m1875/1875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 8ms/step - accuracy: 0.8889 - loss: 0.4049\n", "Epoch 5/5\n", "\u001b[1m1875/1875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 8ms/step - accuracy: 0.8840 - loss: 0.4177\n", "\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - accuracy: 0.9039 - loss: 0.3451\n" ] }, { "data": { "text/plain": [ "[0.30404922366142273, 0.9150999784469604]" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# 官方实例\n", "\n", "import tensorflow as tf\n", "mnist = tf.keras.datasets.mnist\n", "\n", "(x_train, y_train),(x_test, y_test) = mnist.load_data()\n", "x_train, x_test = x_train / 255.0, x_test / 255.0\n", "\n", "model = tf.keras.models.Sequential([\n", " tf.keras.layers.Flatten(input_shape=(28, 28)),\n", " tf.keras.layers.Dense(128, activation='relu'),\n", " tf.keras.layers.Dropout(0.2),\n", " tf.keras.layers.Dense(10, activation='softmax')\n", "])\n", "\n", "model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=0.001),\n", " loss=SparseCategoricalCrossentropy(),\n", " metrics=['accuracy'])\n", "\n", "model.fit(x_train, y_train, epochs=5)\n", "model.evaluate(x_test, y_test)" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/5 - Loss: 0.4415533185513543\n", "Epoch 2/5 - Loss: 0.24540251612599726\n", "Epoch 3/5 - Loss: 0.19663310029716696\n", "Epoch 4/5 - Loss: 0.17048093510557338\n", "Epoch 5/5 - Loss: 0.16070798563876196\n", "Accuracy: 95.38%\n" ] } ], "source": [ "# 用torch实现\n", "\n", "import torch\n", "import torch.nn as nn\n", "import torch.optim as optim\n", "from torchvision import datasets, transforms\n", "\n", "# 数据预处理\n", "transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.5,), (0.5,))])\n", "\n", "# 加载数据集\n", "trainset = datasets.MNIST(root='./data', train=True, download=True, transform=transform)\n", "trainloader = torch.utils.data.DataLoader(trainset, batch_size=64, shuffle=True)\n", "\n", "testset = datasets.MNIST(root='./data', train=False, download=True, transform=transform)\n", "testloader = torch.utils.data.DataLoader(testset, batch_size=64, shuffle=False)\n", "\n", "# 定义模型\n", "class SimpleNet(nn.Module):\n", " def __init__(self):\n", " super(SimpleNet, self).__init__()\n", " self.flatten = nn.Flatten()\n", " self.fc1 = nn.Linear(28 * 28, 128)\n", " self.dropout = nn.Dropout(0.2)\n", " self.fc2 = nn.Linear(128, 10)\n", "\n", " def forward(self, x):\n", " x = self.flatten(x)\n", " x = torch.relu(self.fc1(x))\n", " x = self.dropout(x)\n", " x = self.fc2(x)\n", " return x\n", "\n", "model = SimpleNet()\n", "\n", "if torch.backends.mps.is_available():\n", " device = torch.device(\"mps\")\n", "else:\n", " device = torch.device(\"cpu\")\n", "\n", "model.to(device)\n", "\n", "# 定义损失函数和优化器\n", "criterion = nn.CrossEntropyLoss()\n", "optimizer = optim.Adam(model.parameters(), lr=0.001)\n", "\n", "# 训练模型\n", "epochs = 5\n", "for epoch in range(epochs):\n", " running_loss = 0\n", " for images, labels in trainloader:\n", " images, labels = images.to(device), labels.to(device) # 将数据移动到设备上\n", " optimizer.zero_grad()\n", " output = model(images)\n", " loss = criterion(output, labels)\n", " loss.backward()\n", " optimizer.step()\n", " running_loss += loss.item()\n", " print(f\"Epoch {epoch+1}/{epochs} - Loss: {running_loss/len(trainloader)}\")\n", "\n", "# 测试模型\n", "correct = 0\n", "total = 0\n", "with torch.no_grad():\n", " for images, labels in testloader:\n", " images, labels = images.to(device), labels.to(device) # 将数据移动到设备上\n", " output = model(images)\n", " _, predicted = torch.max(output, 1)\n", " total += labels.size(0)\n", " correct += (predicted == labels).sum().item()\n", "\n", "print(f\"Accuracy: {100 * correct / total}%\")" ] } ], "metadata": { "kernelspec": { "display_name": "ail", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.10.16" } }, "nbformat": 4, "nbformat_minor": 2 }