Pytorch框架班编码-逻辑回归

本次编码的实现环境是Python3.8.3、torch1.5、Anaconda3(64-bit)、PyCharm2020.1。是深度之眼PyTorch框架班的练习及作业，个别代码有修改，仅供交流学习之用。

通过演示逻辑回归模型的训练，学习机器学习回归模型的五大模块：数据、模型、损失函数、优化器和迭代训练过程。

# 逻辑回归模型训练
import os
os.environ['KMP_DUPLICATE_LIB_OK'] = 'True'import torch
import torch.nn as nn
import matplotlib.pyplot as plt
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = Falseimport numpy as np
torch.manual_seed(10)

#-----step1 数据
sample_nums = 100
mean_value = 1.7
bias = 0.5
n_data = torch.ones(sample_nums, 2)
x0 = torch.normal(mean_value * n_data, 1) + bias
y0 = torch.zeros(sample_nums)
x1 = torch.normal(-mean_value * n_data, 1) + bias
y1 = torch.ones(sample_nums)
train_x = torch.cat((x0, x1), 0)
train_y = torch.cat((y0, y1), 0)
max_x = torch.max(train_x)

#-----step2 模型class LR(nn.Module):
    def __init__(self):
        super(LR, self).__init__()
        self.features = nn.Linear(2, 1)
        self.sigmoid = nn.Sigmoid()

    def forward(self, x):
        x = self.features(x)
        x = self.sigmoid(x)
        return x

lr_net = LR()

#-----step3 选择损失函数
loss_fun = nn.BCELoss()

#-----step4 选择优化器
lr = 0.001optimizer = torch.optim.SGD(lr_net.parameters(), lr=lr, momentum=0.9)

#-----step5 模型训练
for iteration in range(1000):
    y_pred = lr_net(train_x)

    loss = loss_fun(y_pred.squeeze(), train_y)

    loss.backward()

    optimizer.step()

    optimizer.zero_grad()

    if iteration % 20 == 0:
        mask = y_pred.ge(0.5).float().squeeze()
        correct = (mask == train_y).sum()
        acc = correct.item() / train_y.size(0)

        plt.scatter(x0.data.numpy()[:, 0], x0.data.numpy()[:, 1], c='r', label='类别0')
        plt.scatter(x1.data.numpy()[:, 0], x1.data.numpy()[:, 1], c='b', label='类别1')

        w0, w1 = lr_net.features.weight[0]
        w0, w1 = float(w0.item()), float(w1.item())
        plot_b = float(lr_net.features.bias[0].item())
        plot_x = np.arange(-6, 6, 0.1)
        plot_y = (-w0 * plot_x - plot_b) / w1

        #plt.xlim(-5, 7)        #plt.ylim(-7, 7)        
        plt.xlim(torch.min(x1[:, 0]), torch.max(x0[:, 0]))
        plt.ylim(torch.min(x1[:, 1]), torch.max(x0[:, 1]))
        plt.plot(plot_x, plot_y)

				plt.text(torch.min(x1[:, 0])-5, torch.max(x0[:, 1])-5, '损失=%.4f' % loss.data.numpy(), fontdict={'size': 20, 'color': 'red'})
        plt.title("迭代次数：{}\nw0:{:.2f} w1:{:.2f} b:{:.2f} 准确率：{:.2f}".format(iteration, w0, w1, plot_b, acc))
        plt.legend()

        plt.show()
        plt.pause(0.5)

        if acc > 0.99:
            break