Black-box optimization benchmarking of IPOP-saACM-ES and BIPOP-saACM-ES on the BBOB-2012 noiseless testbed
This work provides incremental improvements in black-box optimization for benchmarking and algorithm development, showing enhanced performance on specific benchmark functions.
The paper evaluated surrogate-assisted evolution strategies (IPOP-saACM-ES and BIPOP-saACM-ES) on noiseless benchmark problems, finding that they outperformed their surrogate-less versions by a factor of 2 to 4 on 8 out of 24 problems and achieved the best results on specific functions like Ellipsoid and Discus.
In this paper, we study the performance of IPOP-saACM-ES and BIPOP-saACM-ES, recently proposed self-adaptive surrogate-assisted Covariance Matrix Adaptation Evolution Strategies. Both algorithms were tested using restarts till a total number of function evaluations of $10^6D$ was reached, where $D$ is the dimension of the function search space. We compared surrogate-assisted algorithms with their surrogate-less versions IPOP-saACM-ES and BIPOP-saACM-ES, two algorithms with one of the best overall performance observed during the BBOB-2009 and BBOB-2010. The comparison shows that the surrogate-assisted versions outperform the original CMA-ES algorithms by a factor from 2 to 4 on 8 out of 24 noiseless benchmark problems, showing the best results among all algorithms of the BBOB-2009 and BBOB-2010 on Ellipsoid, Discus, Bent Cigar, Sharp Ridge and Sum of different powers functions.