Regression on 42 Regression Datasets (840 train test)
1.037Relative RMSE (50% Quantile)BART-cv
Evaluation results
| Method | Links | ||
|---|---|---|---|
| BART-cvHyperparameter Tuning=5-fold cross-validation2008.06 | 1.037 | 1.117 | |
| Random forest2008.06 | 1.053 | 1.181 | |
| Neural net2008.06 | 1.055 | 1.195 | |
| BART-defaultHyperparameter Tuning=None (off-the-shelf)2008.06 | 1.055 | 1.164 | |
| Boosting2008.06 | 1.068 | 1.189 | |
| Lasso2008.06 | 1.196 | 1.762 |