Black-box optimization benchmarking of IPOP-saACM-ES on the BBOB-2012 noisy testbed
This work provides incremental improvements in black-box optimization benchmarking for researchers in evolutionary computation, focusing on noisy testbeds.
The paper tackled the problem of benchmarking a self-adaptive surrogate-assisted evolution strategy (IPOP-saACM-ES) on noisy optimization functions, showing it outperforms a baseline by a factor of 2 to 3 on 6 problems and exceeds previous records on 15 out of 30 benchmark problems in dimension 20.
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.