### 基于PyTorch实现图像数据6分类后的散点图生成 以下是基于PyTorch框架的一个完整代码示例，该代码实现了对图片数据进行6分类的任务，并在完成后生成散点图。此代码假设输入的数据已经被预处理为适合模型的形式。 #### 数据准备与模型定义 为了简化演示过程，这里仅展示一个基本的卷积神经网络（CNN），并假定已经加载了一个具有6类标签的数据集。 ```python import torch import torch.nn as nn import torch.optim as optim from torchvision import datasets, transforms import matplotlib.pyplot as plt import numpy as np # 定义简单的卷积神经网络 class SimpleCNN(nn.Module): def __init__(self, num_classes=6): super(SimpleCNN, self).__init__() self.conv_layers = nn.Sequential( nn.Conv2d(3, 32, kernel_size=3, stride=1, padding=1), nn.ReLU(), nn.MaxPool2d(kernel_size=2, stride=2), nn.Conv2d(32, 64, kernel_size=3, stride=1, padding=1), nn.ReLU(), nn.MaxPool2d(kernel_size=2, stride=2) ) self.fc_layers = nn.Sequential( nn.Linear(64 * 8 * 8, 128), # 输入大小取决于MaxPool的结果 nn.ReLU(), nn.Linear(128, num_classes) ) def forward(self, x): x = self.conv_layers(x) x = x.view(x.size(0), -1) # 展平 x = self.fc_layers(x) return x # 加载CIFAR-10数据集作为例子 (只取前6类) transform = transforms.Compose([ transforms.Resize((32, 32)), transforms.ToTensor() ]) train_dataset = datasets.CIFAR10(root='./data', train=True, download=True, transform=transform) # 过滤掉不属于前6类别的样本 filtered_indices = [i for i, label in enumerate(train_dataset.targets) if label < 6] subset_train_data = torch.utils.data.Subset(train_dataset, filtered_indices) train_loader = torch.utils.data.DataLoader(subset_train_data, batch_size=64, shuffle=True) device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') model = SimpleCNN(num_classes=6).to(device) criterion = nn.CrossEntropyLoss() optimizer = optim.Adam(model.parameters(), lr=0.001) # 训练模型 num_epochs = 5 for epoch in range(num_epochs): model.train() running_loss = 0.0 for images, labels in train_loader: images, labels = images.to(device), labels.to(device) optimizer.zero_grad() outputs = model(images) loss = criterion(outputs, labels) loss.backward() optimizer.step() running_loss += loss.item() print(f"Epoch [{epoch+1}/{num_epochs}], Loss: {running_loss/len(train_loader):.4f}") ``` #### 散点图生成部分 完成训练后，可以提取特征向量并对它们可视化： ```python def extract_features(dataloader, model): all_features = [] all_labels = [] with torch.no_grad(): model.eval() for images, labels in dataloader: features = model.conv_layers(images.to(device)) features = features.view(features.size(0), -1) # 展平 all_features.append(features.cpu().numpy()) all_labels.extend(labels.numpy()) return np.vstack(all_features), np.array(all_labels) features, true_labels = extract_features(train_loader, model) # 使用PCA降维到二维空间以便绘图 from sklearn.decomposition import PCA pca = PCA(n_components=2) reduced_features = pca.fit_transform(features) plt.figure(figsize=(10, 7)) scatter = plt.scatter(reduced_features[:, 0], reduced_features[:, 1], c=true_labels, cmap='tab10', s=10) plt.colorbar(scatter) plt.title("Scatter Plot of Features after Dimensionality Reduction") plt.xlabel("Principal Component 1") plt.ylabel("Principal Component 2") save_path = "./output_scatter_plot" plt.savefig(f"{save_path}/feature_scatter.png") # 存储散点图 plt.show() ``` 以上代码展示了如何通过PyTorch构建一个简单卷积神经网络来执行6分类任务，并利用PCA技术将高维特征降至两维以生成散点图[^1]。 ---