深度学习实战模拟——softmax回归（图像识别并分类）

1、数据集：

2、完整代码

1、数据集：

1.1 Fashion-MNIST是一个服装分类数据集，由10个类别的图像组成，分别为t-shirt（T恤）、trouser（裤子）、pullover（套衫）、dress（连衣裙）、coat（外套）、sandal（凉鞋）、shirt（衬衫）、sneaker（运动鞋）、bag（包）和ankle boot（短靴）。

1.2 Fashion‐MNIST由10个类别的图像组成，每个类别由训练数据集（train dataset）中的6000张图像和测试数据集（test dataset）中的1000张图像组成。因此，训练集和测试集分别包含60000和10000张图像。测试数据集不会用于训练，只用于评估模型性能。

以下函数用于在数字标签索引及其文本名称之间进行转换。

# 通过ToTensor实例将图像数据从PIL类型变换成32位浮点数格式，
# 并除以255使得所有像素的数值均在0～1之间
trans = transforms.ToTensor()
mnist_train = torchvision.datasets.FashionMNIST(
    root="../data", train=True, transform=trans, download=True)
mnist_test = torchvision.datasets.FashionMNIST(
    root="../data", train=False, transform=trans, download=True)

以下函数用于在数字标签索引及其文本名称之间进行转换。

def get_fashion_mnist_labels(labels):  #@save
    """返回Fashion-MNIST数据集的文本标签"""
    text_labels = ['t-shirt', 'trouser', 'pullover', 'dress', 'coat',
                   'sandal', 'shirt', 'sneaker', 'bag', 'ankle boot']
    return [text_labels[int(i)] for i in labels]

2、完整代码

import torch
import torchvision
import pylab
from torch.utils import data
from torchvision import transforms
import matplotlib.pyplot as plt
from d2l import torch as d2l
import time

batch_size = 256
num_inputs = 784
num_outputs = 10
W = torch.normal(0, 0.01, size=(num_inputs, num_outputs), requires_grad=True)
b = torch.zeros(num_outputs, requires_grad=True)
num_epochs = 5


class Accumulator:
    """在n个变量上累加"""
    def __init__(self, n):
        self.data = [0.0] * n

    def add(self, *args):
        self.data = [a + float(b) for a, b in zip(self.data, args)]

    def reset(self):
        self.data = [0.0] * len(self.data)

    def __getitem__(self, idx):
        return self.data[idx]


def accuracy(y_hat, y):  #@save
    """计算预测正确的数量"""
    if len(y_hat.shape) > 1 and y_hat.shape[1] > 1:
        y_hat = y_hat.argmax(axis=1)
    cmp = y_hat.type(y.dtype) == y
    return float(cmp.type(y.dtype).sum())


def cross_entropy(y_hat, y):
    return -torch.log(y_hat[range(len(y_hat)), y])


def softmax(X):
    X_exp = torch.exp(X)
    partition = X_exp.sum(1, keepdim=True)
    return X_exp/partition


def net(X):
    return softmax(torch.matmul(X.reshape((-1, W.shape[0])), W) + b)


def get_dataloader_workers():
    """使用一个进程来读取的数据"""
    return 1


def get_fashion_mnist_labels(labels):
    """返回Fashion-MNIST数据集的文本标签"""
    #共10个类别
    text_labels = ['t-shirt', 'trouser', 'pullover', 'dress', 'coat', 'sandal', 'shirt', 'sneaker', 'bag', 'ankle boot']
    return [text_labels[int(i)] for i in labels]


def show_images(imgs, num_rows, num_cols, titles=None, scale=1.5):
    """画一系列图片"""
    figsize = (num_cols * scale, num_rows * scale)
    _, axes = plt.subplots(num_rows, num_cols, figsize=figsize)
    for i, (img, label) in enumerate(zip(imgs, titles)):
        xloc, yloc = i//num_cols, i % num_cols
        if torch.is_tensor(img):
            # 图片张量
            axes[xloc, yloc].imshow(img.reshape((28, 28)).numpy())
        else:
            # PIL图片
            axes[xloc, yloc].imshow(img)
        # 设置标题并取消横纵坐标上的刻度
        axes[xloc, yloc].set_title(label)
        plt.xticks([], ())
        axes[xloc, yloc].set_axis_off()
    pylab.show()


def load_data_fashion_mnist(batch_size, resize=None):
    """下载Fashion-MNIST数据集，然后将其加载到内存中"""
    trans = transforms.ToTensor()
    if resize:
        trans.insert(0, transforms.Resize(resize))
    mnist_train = torchvision.datasets.FashionMNIST(root='../data', train=True, transform=trans, download=True)
    mnist_test = torchvision.datasets.FashionMNIST(root='../data', train=False, transform=trans, download=True)
    return (data.DataLoader(mnist_train, batch_size, shuffle=True, num_workers=get_dataloader_workers()),
            data.DataLoader(mnist_test, batch_size, shuffle=False, num_workers=get_dataloader_workers()))


def evaluate_accuracy(net, data_iter):
    """计算在指定数据集上模型的精度"""
    if isinstance(net, torch.nn.Module):
        net.eval()  # 将模型设置为评估模式
    metric = Accumulator(2)  # 正确预测数、预测总数
    with torch.no_grad():
        for X, y in data_iter:
            metric.add(accuracy(net(X), y), y.numel())
    return metric[0] / metric[1]


def updater(batch_size):
    lr = 0.1
    return d2l.sgd([W, b], lr, batch_size)


def train_epoch_ch3(net, train_iter, loss, updater):
    if isinstance(net, torch.nn.Module):
        net.train()
    metric = Accumulator(3)
    for X, y in train_iter:
        y_hat = net(X)
        lo = loss(y_hat, y)
        if isinstance(updater, torch.optim.Optimizer):
            updater.zero_grad()
            lo.backward()
            updater.step()
            metric.add(float(lo)*len(y), accuracy(y_hat, y), y.size().numel())
        else:
            lo.sum().backward()
            updater(X.shape[0])
        metric.add(float(lo.sum()), accuracy(y_hat, y), y.numel())
    return metric[0] / metric[2], metric[1] / metric[2]


class Animator:  #@save
    """绘制数据"""
    def __init__(self, legend=None):
        self.legend = legend
        self.X = [[], [], []]
        self.Y = [[], [], []]

    def add(self, x, y):
        # 向图表中添加多个数据点
        if not hasattr(y, "__len__"):
            y = [y]
        n = len(y)
        if not hasattr(x, "__len__"):
            x = [x] * n
        for i, (a, b) in enumerate(zip(x, y)):
            if a is not None and b is not None:
                self.X[i].append(a)
                self.Y[i].append(b)

    def show(self):
        plt.plot(self.X[0], self.Y[0], 'r--')
        plt.plot(self.X[1], self.Y[1], 'g--')
        plt.plot(self.X[2], self.Y[2], 'b--')
        plt.legend(self.legend)
        plt.xlabel('epoch')
        plt.ylabel('value')
        plt.title('Visual')
        plt.show()


def train_ch3(net, train_iter, test_iter, loss, num_epochs, updater):  #@save
    """训练模型"""
    animator = Animator(legend=['train loss', 'train acc', 'test acc'])
    for epoch in range(num_epochs):
        train_metrics = train_epoch_ch3(net, train_iter, loss, updater)
        train_loss, train_acc = train_metrics
        test_acc = evaluate_accuracy(net, test_iter)
        animator.add(epoch + 1, train_metrics + (test_acc,))
        print(f'epoch: {epoch+1},train_loss:{train_loss:.4f}, train_acc:{train_acc:.4f}, test_acc:{test_acc:.4f}')
    animator.show()


def predict_ch3(net, test_iter, n=12):
    """预测标签"""
    for X, y in test_iter:
        break
    trues = d2l.get_fashion_mnist_labels(y)
    preds = d2l.get_fashion_mnist_labels(net(X).argmax(axis=1))
    titles = [true +'\n' + pred for true, pred in zip(trues, preds)]
    show_images(
        X[0:n].reshape((n, 28, 28)), 2, int(n/2), titles=titles[0:n])


if __name__ == '__main__':
    train_iter, test_iter = load_data_fashion_mnist(batch_size)
    train_ch3(net, train_iter, test_iter, cross_entropy, num_epochs, updater)
    predict_ch3(net, test_iter)

分类效果：