pytorch_搭建简单神经网络（分类）

你的名字 2021-05-12 11:49 851阅读 0赞

*  用pytorch搭建神经网络，将数据分成两类
 *  环境 pytoch1.1.0

### 1. 生成训练数据 ###

# 准备数据
    data = torch.ones(100, 2)
    x1 = torch.normal(mean=2*data, std=1)
    y1 = torch.zeros(100)    # x1 对应的标签
    
    x2 = torch.normal(mean=-2*data, std=1)
    y2 = torch.ones(100)     # x2 对应的标签
    
    train_x = torch.cat((x1, x2), dim=0).type(torch.float32)   # 合并两类数据(x1, x2)，生成训练数据
    label_y = torch.cat((y1, y2), dim=0).type(torch.int64)     # 合并两类便签(y1,y2)，生成真实标签
    
    
    # 通过matplotlib可视化生成的数据
    # plt.scatter(train_x.numpy()[:, 0], train_x.numpy()[:, 1])
    # plt.show()

### 2. 搭建神经网络 ###

# 搭建神经网络
    class Neuro_net(torch.nn.Module):
        """神经网络"""
        def __init__(self, n_feature, n_hidden_layer, n_output):
            super(Neuro_net, self).__init__()
            self.hidden_layer = torch.nn.Linear(n_feature, n_hidden_layer)
            self.output_layer = torch.nn.Linear(n_hidden_layer, n_output)
    
        def forward(self, input):
            hidden_out = torch.relu(self.hidden_layer(input))
            out = self.output_layer(hidden_out)
            return out

### 3. 训练神经网络 ###

num_feature = 2
    num_hidden_layer = 10
    num_output = 2
    epoches = 200
    net = Neuro_net(num_feature, num_hidden_layer, num_output)
    print(net)   # 查看网络结构
    
    # 优化器
    optimizer = torch.optim.SGD(net.parameters(), lr=0.002)
    # 定义损失函数
    loss_function = torch.nn.CrossEntropyLoss()
    
    for epoch in range(epoches):
        out = net(train_x)
        loss = loss_function(out, label_y)
    
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

### 4. 为了更好理解训练过程，加入可视化 ###

import matplotlib.pyplot as plt
    plt.ion()   # 画图
    
    for epoch in range(epoches):
        ...
        loss.backward()
        optimizer.step()
        
            if epoch % 5 == 0:
            print("已训练{}步 | loss：{}。".format(epoch, loss))
            plt.cla()
            pridect_y = torch.max(out, dim=1)[1]
            pridect_label = pridect_y.data.numpy()  # 预测的label
            true_label = label_y.data.numpy()       # 真实的label
    
            plt.scatter(train_x.data.numpy()[:, 0], train_x.data.numpy()[:, 1], c=pridect_label)
            # 计算准确率，显示准确率
            accuracy = float((pridect_label == true_label).astype(int).sum()) / float(true_label.size)
            plt.text(1.5, -4, 'Accuracy=%.2f' % accuracy, fontdict={ 'size': 12, 'color': 'blue'})
            plt.pause(0.1)
    
    plt.ioff()
    plt.show()

### 整体代码： ###

import torch
    import matplotlib.pyplot as plt
    
    
    # 搭建神经网络
    class Neuro_net(torch.nn.Module):
        """神经网络"""
        def __init__(self, n_feature, n_hidden_layer, n_output):
            super(Neuro_net, self).__init__()
            self.hidden_layer = torch.nn.Linear(n_feature, n_hidden_layer)
            self.output_layer = torch.nn.Linear(n_hidden_layer, n_output)
    
        def forward(self, input):
            hidden_out = torch.relu(self.hidden_layer(input))
            out = self.output_layer(hidden_out)
            return out
    
    ## 准备数据
    data = torch.ones(100, 2)
    x1 = torch.normal(mean=2*data, std=1)
    y1 = torch.zeros(100)    # x1 对应的标签
    
    x2 = torch.normal(mean=-2*data, std=1)
    y2 = torch.ones(100)     # x2 对应的标签
    
    train_x = torch.cat((x1, x2), dim=0).type(torch.float32)   # 合并两类数据(x1, x2)，生成训练数据
    label_y = torch.cat((y1, y2), dim=0).type(torch.int64)     # 合并两类便签(y1,y2)，生成真实标签
    
    # # 查看数据分布
    # plt.scatter(train_x.numpy()[:, 0], train_x.numpy()[:, 1])
    # plt.show()
    
    ## 进行训练
    num_feature = 2
    num_hidden_layer = 10
    num_output = 2
    epoches = 200
    net = Neuro_net(num_feature, num_hidden_layer, num_output)
    print(net)   # 查看网络结构
    
    plt.ion()
    # 优化器
    optimizer = torch.optim.SGD(net.parameters(), lr=0.002)
    # 定义损失函数
    loss_function = torch.nn.CrossEntropyLoss()
    
    for epoch in range(epoches):
        out = net(train_x)
        loss = loss_function(out, label_y)
    
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
    
        if epoch % 5 == 0:
            print("已训练{}步 | loss：{}。".format(epoch, loss))
            plt.cla()
            pridect_y = torch.max(out, dim=1)[1]
            pridect_label = pridect_y.data.numpy()  # 预测的label
            true_label = label_y.data.numpy()       # 真实的label
    
            plt.scatter(train_x.data.numpy()[:, 0], train_x.data.numpy()[:, 1], c=pridect_label)
            # 计算准确率，显示准确率
            accuracy = float((pridect_label == true_label).astype(int).sum()) / float(true_label.size)
            plt.text(1.5, -4, 'Accuracy=%.2f' % accuracy, fontdict={ 'size': 12, 'color': 'blue'})
            plt.pause(0.1)
    
    plt.ioff()
    plt.show()

### 可视化动态拟合图 ###

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDkxMjE1OQ_size_16_color_FFFFFF_t_70]

[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDkxMjE1OQ_size_16_color_FFFFFF_t_70]: /images/20210511/8f32ede10bf4486f9cdc6765de14c8a6.png