Black-box optimization benchmarking the IPOP-CMA-ES on the noiseless testbed: comparison to the BIPOP-CMA-ES

We benchmark the Covariance Matrix Adaptation-Evolution Strategy (CMA-ES) algorithm with an Increasing POPulation size (IPOP) restart policy on the BBOB noiseless testbed. The IPOP-CMA-ES is compared to the BIPOPCMA-ES and is shown to perform at best two times faster on multi-modal functions f15 to f19 whereas it does not solve weakly structured functions f22, f23 and f24. Categories and Subject Descriptors

• Type: Journal
• Year: 2010
• Where: GECCO
• Authors: Raymond Ros
• Keywords: Covariance Matrix Adaptation-Evolution, GECCO 2010, Multi-modal Functions, Optimization, Subject Descriptors