Table 5:
Parameters for feature subsets
| Parameter(s) | Tested values |
|---|---|
| Learning rate | 0.01, 0.04 |
| Maximum tree depth (max_depth) | 3 |
| Data subsampling ratio (subsample) | 0.8, 1 |
| Minimum weight of a leaf node (min_child_weight) | 2, 4 |
| L1 regularization (alpha) | 1, 2 |
| L2 regularization (lambda) | 0 |
| Number of trees to fit per iteration (n_estimators) | 1 |