Comparison with scikit-optimize
Note
在开始对比测试前，需安装scikit-optimize
Synthetic Function Optimization
测试函数：Branin (2d)
测试优化器：Bayesian Optimization
初始设计：Sobol design
Budget: 200
Repeat num: 10
| Method | Minimum | Best minimum | Mean f_calls to min | Std f_calls to min | Fastest f_calls to min |
|---|---|---|---|---|---|
| skopt(gp_minimize) | 0.398+/-0.000 | 0.398 | 24.4 | 2.69072 | 22 |
| skopt(forest_minimize) | 0.417+/-0.033 | 0.398 | 118.1 | 36.4786 | 31 |
| skopt(gbrt_minimize) | 0.550+/-0.196 | 0.399 | 132.0 | 61.236 | 11 |
| skopt(dummy_minimize) | 0.662+/-0.167 | 0.483 | 86.1 | 54.769 | 19 |
| XBBO(bo-gp) | 0.398+/-0.000 | 0.398 | 42.0 | 5.0398 | 30 |