训练ANN时出现IndexError: Target 5 is out of bounds问题求助

问题描述 #

训练人工神经网络时遇到以下错误：

IndexError: Target 5 is out of bounds

怀疑问题出在数据拆分代码：

from sklearn.datasets import fetch_california_housing
california = fetch_california_housing()
data = pd.DataFrame(california.data)
data.columns = california.feature_names
data['Price'] = california.target

X = data.iloc[:, 0:8]
y = data.iloc[:, 8]

完整错误栈：

IndexError                                Traceback (most recent call last)
Input In [174], in <cell line: 58>()
     55     plt.title("California House Prices Training Loss")
     56     plt.show()
---> 58 train(classifier, optimizer, epochs, loss_fn)

Input In [174], in train(classifier, optimizer, epochs, loss_fn)
     43 for epoch in range(epochs):
     44     out = classifier(X_train)
---> 45     loss = loss_fn(out, y_train)
     46     loss.backward()
     47     optimizer.step()

File ~/opt/anaconda3/lib/python3.9/site-packages/torch/nn/modules/module.py:1194, in Module._call_impl(self, *input, **kwargs)
   1190 # If we don't have any hooks, we want to skip the rest of the logic in
   1191 # this function, and just call forward.
   1192 if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks
   1193         or _global_forward_hooks or _global_forward_pre_hooks):
-> 1194     return forward_call(*input, **kwargs)
   1195 # Do not call functions when jit is used
   1196 full_backward_hooks, non_full_backward_hooks = [], []

File ~/opt/anaconda3/lib/python3.9/site-packages/torch/nn/modules/loss.py:1174, in CrossEntropyLoss.forward(self, input, target)
   1173 def forward(self, input: Tensor, target: Tensor) -> Tensor:
-> 1174     return F.cross_entropy(input, target, weight=self.weight,
   1175                            ignore_index=self.ignore_index, reduction=self.reduction,
   1176                            label_smoothing=self.label_smoothing)

File ~/opt/anaconda3/lib/python3.9/site-packages/torch/nn/functional.py:3026, in cross_entropy(input, target, weight, size_average, ignore_index, reduce, reduction, label_smoothing)
   3024 if size_average is not None or reduce is not None:
   3025     reduction = _Reduction.legacy_get_string(size_average, reduce)
-> 3026 return torch._C._nn.cross_entropy_loss(input, target, weight, _Reduction.get_enum(reduction), ignore_index, label_smoothing)

IndexError: Target 5 is out of bounds.

问题诊断 #

你提供的数据拆分代码没有问题，错误核心原因是：

• 加州房价数据集是回归任务（目标变量Price是连续的房价数值），但你误用了仅适用于分类任务的CrossEntropyLoss损失函数。
• CrossEntropyLoss要求目标值是类别索引（如0、1、2...），且索引值不能超过模型输出的类别数减一。而你的目标变量y是连续的房价（可能出现5左右的数值），损失函数误将其当作分类标签，该数值超出模型输出的类别范围，从而抛出索引越界错误。

修复方案 #

1. 替换损失函数 #

将分类用的CrossEntropyLoss换成回归任务专用的损失函数，比如：

# 均方误差损失（回归任务最常用）
loss_fn = torch.nn.MSELoss()

# 或者平均绝对误差损失
# loss_fn = torch.nn.L1Loss()

2. 调整模型输出层 #

回归任务的输出层不需要Softmax激活，直接输出单个连续值即可（输出维度设为1），示例模型定义：

import torch.nn as nn

class HousePriceRegressor(nn.Module):
    def __init__(self):
        super().__init__()
        self.network = nn.Sequential(
            nn.Linear(8, 64),  # 输入维度为8（数据集特征数）
            nn.ReLU(),
            nn.Linear(64, 32),
            nn.ReLU(),
            nn.Linear(32, 1)   # 回归任务输出维度为1
        )
    
    def forward(self, x):
        return self.network(x)

补充说明 #

加州房价数据集的目标变量Price取值范围约为0.15到5万美元，属于典型的回归任务，必须匹配回归类损失函数和对应模型结构，不能混用分类任务的组件。

内容的提问来源于stack exchange，提问作者George Garman