PyTorch 深度学习实践第六讲（逻辑回归）

Logistic Function

Sigmoid Functions

交叉熵

备注：BCE越小分布越接近，越好

深度学习基本框架

上课代码

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

x_data = torch.Tensor([[1.0],[2.0],[3.0]])
y_data = torch.Tensor([[0],[0],[1]])    #分别表示分类

class LinearModel(torch.nn.Module):
    def __init__(self):
        super(LinearModel,self).__init__()
        self.linear = torch.nn.Linear(1,1)
    def forward(self,x):
        y_pred = F.sigmoid(self.linear(x))     #加入逻辑变换
        return y_pred 
    
model = LinearModel()

criterion = torch.nn.BCELoss(size_average = False)   #将原来MSE变为BCE
optimizer = torch.optim.SGD(model.parameters(),lr = 0.01)   #优化器

for epoch in range(1000):   #training cycle
    y_pred = model(x_data)
    loss = criterion(y_pred,y_data)
    print(epoch,loss.item())
    
    optimizer.zero_grad()
    loss.backward()
    optimizer.step()
    
x = np.linspace(0,10,200)
x_t = torch.Tensor(x).view(200,1)   #转化成200行，1列的矩阵
y_t = model(x_t)
y = y_t.data.numpy()

plt.plot(x,y)
plt.plot([0,10],[0.5,0.5],c = 'c')
plt.xlabel('Hours')
plt.ylabel('probablility of pass')
plt.grid()
plt.show()