Black-box optimization benchmarking of IPOP-saACM-ES on the BBOB-2012 noisy testbed

TL;DR

This paper evaluates the performance of the IPOP-saACM-ES algorithm on the noisy BBOB-2012 benchmark, demonstrating its robustness and superior performance over previous methods on several benchmark problems.

Contribution

It provides the first comprehensive benchmarking of IPOP-saACM-ES on the BBOB-2012 noisy testbed, showing its effectiveness and improvements over prior algorithms.

Findings

IPOP-saACM-ES is as robust as IPOP-aCMA-ES.

It outperforms IPOP-aCMA-ES by 2 to 3 times on 6 benchmark problems.

Exceeds previous records on 15 out of 30 problems in dimension 20.

Abstract

In this paper, we study the performance of IPOP-saACM-ES, recently proposed self-adaptive surrogate-assisted Covariance Matrix Adaptation Evolution Strategy. The algorithm was 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. The experiments show that the surrogate model control allows IPOP-saACM-ES to be as robust as the original IPOP-aCMA-ES and outperforms the latter by a factor from 2 to 3 on 6 benchmark problems with moderate noise. On 15 out of 30 benchmark problems in dimension 20, IPOP-saACM-ES exceeds the records observed during BBOB-2009 and BBOB-2010.