Table 3 Result of determining the best parameters in GBM model based on PSO algorithm.
From: GIS-based non-grain cultivated land susceptibility prediction using data mining methods
Parameters | GBM | PSO-GBM |
|---|---|---|
n-trees | 300 | 437 |
Max-features | 1 | 0.427906 |
Min-samples-split | 3 | 2 |
Min-samples-leaf | 1 | 3 |
Tree depth | 4 | 6 |
Learning rate | 0.100000 | 0.287674 |
