深度学习框架Pytorch——学习笔记(四)逻辑回归
在上一个线性回归的基础上进行实现逻辑回归
逻辑回归中用到了sigmoid函数
可以理解为将结果值转为概率问题(0,1)之间,逻辑回归实际上就是一个二分类了,将结果分为 0 and 1 .
import torch
from torch.autograd import Variable
# 获取数据
x_data = Variable(torch.Tensor([[0.6], [1.0], [3.5], [4.0]]))
y_data = Variable(torch.Tensor([[0.], [0.], [1.], [1.]]))
class Model(torch.nn.Module):
def __init__(self):
super(Model, self).__init__()
self.linear = torch.nn.Linear(1, 1)
self.sigmoid = torch.nn.Sigmoid()
def forward(self, x):
y_pred = self.sigmoid(self.linear(x))
return y_pred
model = Model()
# 定义损失函数和优化器
criterion = torch.nn.BCELoss(size_average=True)
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)
# Training
for epoch in range(10000):
# forward
y_pred = model(x_data)
loss = criterion(y_pred, y_data)
if epoch % 20 == 0:
print(epoch, loss.data.numpy())
# Zero gradients
optimizer.zero_grad()
# Backward pass
loss.backward()
# update weights
optimizer.step()
# 测试结果
model.eval()
hour_var = Variable(torch.Tensor([[0.5]]))
print("predict (after training)", 0.5, model(hour_var).data[0][0].numpy())
hour_var = Variable(torch.Tensor([[7.0]]))
print("predict (after training)", 7.0, model(hour_var).data[0][0].numpy())
总结
1. 在线性回归的基础上,比较容易理解逻辑回归。
2. 发现多写代码更容易理解其思路