AcidMap/app/models/cd_prediction/effective_cd/constrained_nn6C.py

import torch
import torch.nn as nn

# 定义神经网络模型
class ConstrainedNN(nn.Module):
    def __init__(self, input_size):
        super(ConstrainedNN, self).__init__()
        self.fc1 = nn.Linear(input_size, 128)
        self.fc2 = nn.Linear(128, 64)
        self.fc3 = nn.Linear(64, 32)
#        self.dropout = nn.Dropout(0.5)
        self.fc4 = nn.Linear(32, 1)
        # 嵌入一元一次线性函数的参数
        self.a = nn.Parameter(torch.randn(1))
        self.b = nn.Parameter(torch.randn(1))

    def forward(self, x, original_x, special_feature_index):
        x = torch.relu(self.fc1(x))
        x = torch.relu(self.fc2(x))
        x = torch.relu(self.fc3(x))
        x = self.fc4(x)
        # 嵌入一元一次线性函数
        special_feature = original_x[:, special_feature_index].unsqueeze(1)
        x = x + self.a * special_feature + self.b
        return x

# 自定义损失函数
def custom_loss(y_pred, y_true, X, monotonic_feature_index, mean_X_special, mean_y):
    y_true = y_true
    mse_loss = nn.MSELoss()(y_pred, y_true)
    # 惩罚项
    X_special = X[:, monotonic_feature_index]
    punishment = torch.mean((torch.sign(X_special - mean_X_special) - torch.sign(y_pred - mean_y)) ** 2)
    return mse_loss + 2*punishment  # 权衡 mse_loss 和惩罚项的权重