MVS experiments in lightgbm hyperparameters for baseline, datasets and data preprocessing are taken from https://github.com/catboost/benchmarks/tree/master/quality_benchmarks Hyperamaters for MVS and GOSS are taken from https://arxiv.org/abs/1910.13204 Summary table for relative error change for different subsampling strategies (columns subsample size):
| 0.05 | 0.1625 | 0.275 | 0.3875 | 0.5 | 0.6125 | 0.725 | 0.8375 | 0.95 |
|---|---|---|---|---|---|---|---|---|
| mvs | 8.314 | 3.032 | 0.775 | 0.061 | -0.391 | -0.403 | -0.26 | -0.185 |
| sgb | 13.037 | 4.516 | 3.169 | 2.569 | 2.105 | 1.521 | 1.08 | 0.343 |
| goss | 17.263 | 4.672 | 3.285 | 2.223 | 1.196 | 0.488 | 0.007 | -0.05 |
| mvs_adaptive | 8.314 | 2.919 | 0.904 | -0.195 | -0.451 | -0.466 | -0.274 | -0.079 |