House Price Prediction on California Housing 1990 U.S. Census (train)
0.2771RMSHuber Loss
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Huber LossLoss Function=Huber Loss, Model=Feedforward Network, Optimizer=Adam, Batch Size=32, Epochs=30, Trials=50, Preprocessing=Min-max normalization, Add 0.5 noise to 20% of the training labels2026.06 | 0.2771 | 0.2242 | 0.0888 | |
| Log CoshLoss Function=Log Cosh, Model=Feedforward Network, Optimizer=Adam, Batch Size=32, Epochs=30, Trials=50, Preprocessing=Min-max normalization, Add 0.5 noise to 20% of the training labels2026.06 | 0.2772 | 0.2241 | 0.0881 | |
| MSELoss Function=MSE, Model=Feedforward Network, Optimizer=Adam, Batch Size=32, Epochs=30, Trials=50, Preprocessing=Min-max normalization, Add 0.5 noise to 20% of the training labels2026.06 | 0.2838 | 0.2316 | 0.0303 | |
| SMAELoss Function=SMAE, Model=Feedforward Network, Optimizer=Adam, Batch Size=32, Epochs=30, Trials=50, Preprocessing=Min-max normalization, Add 0.5 noise to 20% of the training labels2026.06 | 0.2839 | 0.2273 | 0.0297 | |
| SRLLoss Function=SRL, Model=Feedforward Network, Optimizer=Adam, Batch Size=32, Epochs=30, Trials=50, Preprocessing=Min-max normalization, Add 0.5 noise to 20% of the training labels2026.06 | 0.3017 | 0.2362 | -0.0899 |