pytorch解决分类问题神经网络

 

目录

 

1.pycharm动态显示代码

2.pycharm结果

 3.jupyter notebook 动态显示代码

4.jupyter notebook显示结果

---

1.pycharm动态显示代码

import matplotlib.pyplot as plt
import torch
import torch.nn as nn
data=torch.ones(100,2)
x0=torch.normal(2*data,1)#(均值，方差）
y0=torch.zeros(100)#第一类标签设置为0
x1=torch.normal(-2*data,1)
y1=torch.ones(100)#第一类标签设置为1
x=torch.cat((x0,x1)).type(torch.FloatTensor)#x1,x2合并成x并转化成Float
y=torch.cat((y0,y1)).type(torch.LongTensor)

class Net(nn.Module):
    def __init__(self):
        super(Net,self).__init__()
        self.classify=nn.Sequential(
            nn.Linear(2,15),
            nn.ReLU(),
            nn.Linear(15,2),
            nn.Softmax(dim=1)

        )
    def forward(self,x):
        classification=self.classify(x)
        return classification

net=Net()
optimizer=torch.optim.SGD(net.parameters(),lr=0.03)
loss_func=nn.CrossEntropyLoss()
plt.ion()
for epoch in range(100):
    out=net(x)
    loss=loss_func(out,y)
    optimizer.zero_grad()
    loss.backward()
    optimizer.step()
    if epoch%2==0:
        plt.cla()
        classification=torch.max(out,1)[1]#返回每一行中最大值的下标
        class_y=classification.data.numpy()
        taget_y=y.data.numpy()
        plt.scatter(x.data.numpy()[:,0],x.data.numpy()[:,1],c=class_y,s=100,cmap="RdYlGn")#c=class_y表示颜色随着class_y的值不同而变化，颜色模式采用"RdYlGn"

        accuracy=sum(class_y==taget_y)/200
        plt.text(1.5,-4,f'Accuracy={accuracy}',fontdict={'size':20,'color':'red'})
        plt.pause(0.4)
    plt.show()
plt.ioff()
plt.show()

2.pycharm结果

 3.jupyter notebook 动态显示代码

与pycharm中代码整体类似，只是不用打开和关闭交互模式，使用display库中函数清空已经画的图像

from IPython import display

for epoch in range(100):
    out=net(x)
    loss=loss_func(out,y)
    optimizer.zero_grad()
    loss.backward()
    optimizer.step()
    if epoch%2==0:
        display.clear_output(wait=True)
        classification=torch.max(out,1)[1]#返回每一行中最大值的下标
        class_y=classification.data.numpy()
        taget_y=y.data.numpy()
        plt.scatter(x.data.numpy()[:,0],x.data.numpy()[:,1],c=class_y,s=100,cmap="RdYlGn")#c=class_y表示颜色随着class_y的值不同而变化，颜色模式采用"RdYlGn"

        accuracy=sum(class_y==taget_y)/200
        plt.text(1.5,-4,f'Accuracy={accuracy}',fontdict={'size':20,'color':'red'})
        plt.pause(0.3)
    plt.show()

4.jupyter notebook显示结果