Skip to Main Content
Table 9:
Performance of the algorithm selection model on the test set, where is the percentage of instances for which an algorithm is best and % is the model forecast. The remaining columns represent the confusion matrix of the model. In boldface are those values in the diagonal above 90%, which represent excellent model performance, and off the diagonal above 10%, which represent poor model performance.
BIPOP-
Algorithm%BFGSCMA-ESLSstepNelder–DoerrCMA-ES
BFGS 15.6% 14.0% 77.3% 0.5% 0.0% 3.3% 11.4% 
BIPOP-CMA-ES 42.7% 45.0% 5.3% 92.2% 26.7% 6.7% 8.6% 
LSstep 9.4% 8.3% 0.0% 2.9% 66.7% 3.3% 0.0% 
Nelder–Doerr 25.0% 25.8% 10.7% 3.9% 6.7% 86.7% 2.9% 
CMA-ES 7.3% 6.9% 6.7% 0.5% 0.0% 0.0% 77.1% 
BIPOP-
Algorithm%BFGSCMA-ESLSstepNelder–DoerrCMA-ES
BFGS 15.6% 14.0% 77.3% 0.5% 0.0% 3.3% 11.4% 
BIPOP-CMA-ES 42.7% 45.0% 5.3% 92.2% 26.7% 6.7% 8.6% 
LSstep 9.4% 8.3% 0.0% 2.9% 66.7% 3.3% 0.0% 
Nelder–Doerr 25.0% 25.8% 10.7% 3.9% 6.7% 86.7% 2.9% 
CMA-ES 7.3% 6.9% 6.7% 0.5% 0.0% 0.0% 77.1% 
Close Modal

or Create an Account

Close Modal
Close Modal