Skip to Main Content
Table 4:
Performance of the algorithm performance model on the test set, where is the percentage of solvable instances and is the model forecast. The remaining columns measure the performance of the model. In boldface are values above 90% for ACC, SPC, TPR, NPV, and PPV, representing good model performance, and values above 10% of FNR and FPR, representing poor model performance.
Algorithm%%ACCSPCFNRFPRTPRNPVPPV
BFGS 43.8% 43.8% 98.8% 98.6% 1.1% 1.4% 98.9% 98.6% 98.9% 
BIPOP-CMA-ES 95.8% 95.6% 99.4% 99.6% 5.0% 0.4% 95.0% 99.8% 90.5% 
LSstep 30.2% 31.0% 96.3% 95.2% 3.3% 4.8% 96.7% 92.6% 97.9% 
Nelder–Doerr 59.4% 58.3% 93.1% 93.3% 7.2% 6.7% 92.8% 95.0% 90.5% 
CMA-ES 65.6% 62.5% 90.2% 90.2% 9.7% 9.8% 90.3% 94.7% 82.8% 
Algorithm%%ACCSPCFNRFPRTPRNPVPPV
BFGS 43.8% 43.8% 98.8% 98.6% 1.1% 1.4% 98.9% 98.6% 98.9% 
BIPOP-CMA-ES 95.8% 95.6% 99.4% 99.6% 5.0% 0.4% 95.0% 99.8% 90.5% 
LSstep 30.2% 31.0% 96.3% 95.2% 3.3% 4.8% 96.7% 92.6% 97.9% 
Nelder–Doerr 59.4% 58.3% 93.1% 93.3% 7.2% 6.7% 92.8% 95.0% 90.5% 
CMA-ES 65.6% 62.5% 90.2% 90.2% 9.7% 9.8% 90.3% 94.7% 82.8% 
Close Modal

or Create an Account

Close Modal
Close Modal