PyTorch實現Logistic Regression

1.PyTorch基礎實現Logistic regression

import torch
from torch.autograd import Variable
 
torch.manual_seed(2)
x_data = Variable(torch.Tensor([[1.0], [2.0], [3.0], [4.0]]))
y_data = Variable(torch.Tensor([[0.0], [0.0], [1.0], [1.0]]))
 
# 初始化
w = Variable(torch.Tensor([-1]), requires_grad=True)
b = Variable(torch.Tensor([0]), requires_grad=True)
 
epochs = 100
costs = []
lr = 0.1
 
print('before training,predict of x = 1.5 is :')
print('Y_pred = ', float(w.data * 1.5 + b.data > 0))

 
# 模型訓練
for epoch in range(epochs):
    # 計算梯度
    A = 1 / (1 + torch.exp(-(w * x_data + b)))
    # 邏輯損失函數
    J = - torch.mean(y_data * torch.log(A) + (1 - y_data) * torch.log(1 - A))
    # 自動反向傳播
    J.backward()
 
    # 參數更新
    w.data = w.data - lr * w.grad.data
    w.grad.data.zero_()
    b.data = b.data - lr * b.grad.data
    b.grad.data.zero_()
 
# 模型測試
print('after trainning,predict of x = 1.5 is :')
print('Y_pred = ', float(w.data * 1.5 + b.data > 0))
print(w.data, b.data)

2. 用PyTorch類實現Logistic regression,torch.nn.module寫網絡結構

import torch
# from torch import nn
# 第一創建數據
from torch.autograd import Variable  # 導入Variable函數進行自動求導，有了Variable PyTorch才能實現自動求導功能
 
torch.manual_seed(2)
x_data = Variable(torch.Tensor([[1.0], [2.0], [3.0], [4.0]]))
y_data = Variable(torch.Tensor([[0.0], [0.0], [1.0], [1.0]]))
 
 
# 定義網絡模型
# 先建立一個基類Model，都是從父類torch.nn.Module中繼承過來，PyTorch寫網絡的固定寫法
class Model(torch.nn.Module):
    def __init__(self):
        super(Model, self).__init__()  # 初始父類
        self.linear = torch.nn.Linear(1, 1)  # 輸入維度和輸出維度都爲1
 
    def forward(self, x):
        y_pred = self.linear(x)
        return y_pred
 
model = Model()  # 實例化
 
# 定義Loss和優化方法
criterion = torch.nn.BCEWithLogitsLoss()  # 損失函數，封裝好的邏輯損失函數
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)  # 進行優化梯度下降
# before training
hour_var = Variable(torch.Tensor([[2.5]]))
y_pred = model(hour_var)
print("predict (before training) given", 4, "is", float(model(hour_var).data[0][0] > 0.5))
 
epochs = 40
for epoch in range(epochs):
    # 計算grads and cost
    y_pred = model(x_data)  # x_data 輸入數據進入模型中
    loss = criterion(y_pred, y_data)
    # print(loss.data)
    optimizer.zero_grad()  # 梯度清零
    loss.backward()  # 反向傳播
    optimizer.step()  # 優化迭代
 
# after trining
hour_var = Variable(torch.Tensor([[4.0]]))
y_pred = model(hour_var)  # 預測結果
print("predict (after training) given", 4, "is", float(model(hour_var).data[0][0] > 0.5))