Table 7 Mean absolute error (MAE) comparison across dataset sizes.
From: Scalable earthquake magnitude prediction using spatio-temporal data and model versioning
Algorithm | MAE (100%) | MAE (65%) | MAE (35%) | MAE (20%) | MAE Change (%) |
|---|---|---|---|---|---|
LightGBM | 0.4029 | 0.4018 | 0.4034 | 0.4450 | 10.46 |
XGBoost | 0.4323 | 0.4350 | 0.4319 | 0.4515 | 4.44 |
Gradient Boosting | 0.4081 | 0.4036 | 0.3970 | 0.4323 | 5.92 |
Random Forest | 0.4161 | 0.4220 | 0.4193 | 0.4509 | 8.37 |
Linear Regression | 0.5393 | 0.5347 | 0.5387 | 0.5385 | 0.15 |
