一个简单的基于PyTorch的多层感知器（MLP）神经网络

构建一个简单的神经网络使用PyTorch同样简单。以下是使用PyTorch构建一个多层感知器（MLP）进行分类任务的步骤，我们仍然以鸢尾花（Iris）数据集为例：

1. 导入所需的库

import torch
import torch.nn as nn
import torch.optim as optim
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import OneHotEncoder
from torch.utils.data import TensorDataset, DataLoader

2. 加载和预处理数据

# 加载数据集
iris = load_iris()
X = iris.data
y = iris.target

# 将标签转换为One-hot编码
encoder = OneHotEncoder(sparse=False)
y = encoder.fit_transform(y.reshape(-1, 1))

# 分割数据集为训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# 转换为PyTorch张量
X_train = torch.tensor(X_train, dtype=torch.float32)
y_train = torch.tensor(y_train, dtype=torch.float32)
X_test = torch.tensor(X_test, dtype=torch.float32)
y_test = torch.tensor(y_test, dtype=torch.float32)

# 创建数据加载器
train_data = TensorDataset(X_train, y_train)
test_data = TensorDataset(X_test, y_test)

train_loader = DataLoader(train_data, batch_size=5, shuffle=True)
test_loader = DataLoader(test_data, batch_size=5, shuffle=False)

3. 构建模型

class MLP(nn.Module):
    def __init__(self):
        super(MLP, self).__init__()
        self.layers = nn.Sequential(
            nn.Linear(4, 10),  # 输入层，4个特征到10个节点
            nn.ReLU(),         # 激活函数
            nn.Linear(10, 10), # 隐藏层
            nn.ReLU(),         # 激活函数
            nn.Linear(10, 3)   # 输出层，3个类别
        )

    def forward(self, x):
        return self.layers(x)

model = MLP()

4. 定义损失函数和优化器

criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(), lr=0.01)

5. 训练模型

# 训练模型
epochs = 100
for epoch in range(epochs):
    for batch_features, batch_labels in train_loader:
        optimizer.zero_grad()
        output = model(batch_features)
        loss = criterion(output, torch.argmax(batch_labels, dim=1))
        loss.backward()
        optimizer.step()
    print(f'Epoch {epoch+1}/{epochs}, Loss: {loss.item()}')

6. 评估模型

# 在测试集上评估模型
model.eval()
with torch.no_grad():
    correct = 0
    total = 0
    for batch_features, batch_labels in test_loader:
        outputs = model(batch_features)
        _, predicted = torch.max(outputs.data, 1)
        total += batch_labels.size(0)
        correct += (predicted == torch.argmax(batch_labels, dim=1)).sum().item()

accuracy = 100 * correct / total
print(f'Test accuracy: {accuracy}%')

以上代码展示了如何使用PyTorch构建一个简单的神经网络，并在鸢尾花数据集上进行训练和评估。请注意，为了运行此代码，您需要安装PyTorch和相关库。