PyTorch深度学习实践（06.Logistic Regression）

源自课程：《PyTorch深度学习实践》完结合集

Chapter6 逻辑斯蒂回归

"""
使用logistic回归实现非线性拟合
"""

import torch
import torch.nn.functional as F
import numpy as np
import matplotlib.pyplot as plt

import warnings

warnings.filterwarnings("ignore")

x_data = torch.Tensor([[1.0], [2.0], [3.0]])
y_data = torch.Tensor([[0], [0], [1]])


# 构建模型、将原有的线性模型替换为logistic回归模型
class LogisticRegressionModel(torch.nn.Module):
    def __init__(self):
        super(LogisticRegressionModel, self).__init__()
        self.linear = torch.nn.Linear(1, 1)

    def forwrad(self, x):
        y_pred = F.sigmoid(self.linear(x))
        return y_pred


# 将模型进行实例化
model = LogisticRegressionModel()

# 构建损失函数和优化器
criterion = torch.nn.BCELoss(size_average=False)
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)


# 开始进行迭代
for epoch in range(600):
    # 前向传播、计算损失、打印训练过程损失值
    y_pred = model.forwrad(x_data)
    loss = criterion(y_pred, y_data)
    print("\tloss: ", epoch, loss.item())

    # 梯度清零、反向传播、参数更新
    optimizer.zero_grad()
    loss.backward()
    optimizer.step()

# 构造数据进行测试
x = np.linspace(0, 10, 200)
x_t = torch.Tensor(x).view((200, 1))
y_t = model.forwrad(x_t)
y = y_t.data.numpy()

# 画图
plt.plot(x, y)
plt.plot([0, 10], [0.5, 0.5], c='r')
plt.xlabel('Hours')
plt.ylabel('Probability of Pass')
plt.grid()
plt.show()