模型的选择、过拟合与欠拟合

导入包和模块

%matplotlib inline
import torch
import numpy as np
import sys
sys.path.append("..")
import d2lzh_pytorch as d2l

生成数据集

n_train, n_test, true_w, true_b = 100, 100, [1.2, -3.4, 5.6], 5
# 训练集和测试集数据样本设为100，权重w=[1.2, -3.4 5.6],偏差b为 5
features = torch.randn((n_train + n_test, 1))
poly_features = torch.cat((features, torch.pow(features, 2),
                          torch.pow(features, 3)), 1)
labels = (true_w[0] * poly_features[:, 0]
         + true_w[1] * poly_features[:, 1]
         + true_w[2] * poly_features[:, 2] + true_b)
labels += torch.tensor(np.random.normal(0, 0.1, size=labels.size()),
                      dtype=torch.float)

查看数据特征和分类标签

features[:2], poly_features[:2], labels[:2]

定义，训练和测试模型

def semilogy(x_vals, y_vals, x_label, y_label,
           x2_vals=None, y2_vals=None,
           legend=None, figsize=(3.5, 2.5)):
    d2l.set_figsize(figsize)
    d2l.plt.xlabel(x_label)
    d2l.plt.ylabel(y_label)
    d2l.plt.semilogy(x_vals, y_vals)
    if x2_vals and y2_vals:
        d2l.plt.semilogy(x2_vals, y2_vals, linestyle=':')
        d2l.plt.legend(legend)

num_epochs, loss = 100, torch.nn.MSELoss()

# 初始化网络模型
def fit_and_plot(train_features, test_features,
                train_labels, test_labels):
    net = torch.nn.Linear(train_features.shape[-1], 1)
    
    # 设置批量大小
    batch_size = min(10, train_labels.shape[0])
    dataset = torch.utils.data.TensorDataset(train_features,
                                             train_labels)  # 设置数据集
    train_iter = torch.utils.data.DataLoader(dataset,
                                            batch_size,
                                             shuffle=True)  # 设置获取数据方式
    # 设置优化函数，使用的是随机梯度下降优化
    optimizer = torch.optim.SGD(net.parameters(), lr=0.01)
    
    train_ls, test_ls = [], []
    for _ in range(num_epochs):
        for X, y in train_iter:  # 取一个批量的数据
            # 输入到网络中计算输出，并和标签比较求得损失函数
            l = loss(net(X), y.view(-1, 1))  
            optimizer.zero_grad()  # 梯度清零，防止梯度累加干扰优化
            l.backward()  # 求梯度
            optimizer.step()  # 迭代优化函数，进行参数优化
        train_labels = train_labels.view(-1, 1)
        test_labels = test_labels.view(-1, 1)
        train_ls.append(loss(net(train_features), train_labels).item())# 将训练损失保存到train_ls中
        test_ls.append(loss(net(test_features), test_labels).item())  # 将测试损失保存到test_ls中
        
    print('final epoch: train_loss', train_ls[-1], 'test loss', test_ls[-1])
    semilogy(range(1, num_epochs + 1), train_ls, 'epochs','loss',
            range(1, num_epochs + 1), test_ls, ['train', 'test'])
    print('weight:', net.weight.data,
         '\nbias:',net.bias.data)

三阶多项式函数拟合

fit_and_plot(poly_features[:n_train, :], poly_features[n_train:, :],
            labels[:n_train], labels[n_train:])
# 参数接近真实参数

线性函数拟合（欠拟合）

fit_and_plot(features[:n_train, :], features[n_train:, :],
            labels[:n_train], labels[n_train:])
# 误差较大

训练样本不足（过拟合）

fit_and_plot(poly_features[0:2, :], poly_features[n_train:, :],
            labels[0:2], labels[n_train:])
# 随着迭代次数增加，模型在训练集上的误差越来越小，但是在测试集上误差非常大