Evaluating the Population Size Adaptation Mechanism for CMA-ES on the BBOB Noiseless Testbed
Kouhei Nishida, Youhei Akimoto · 2016
The population size adaptation mechanism for CMA-ES is evaluated on the BBOB noiseless testbed. The population size is adapted on the basis of the estimated accuracy of the update of the distribution parameters, i.e., the mean vector and the covariance matrix of the Gaussian distribution. The population size is adapted so that the estimated accuracy of the parameter update keeps a certain level. The CMA- ES with the population size adaptation mechanism could solve well-structured multimodal functions as efficiently as the best 2009 portfolio without a restart strategy that in- creases the population size every restart such as the IPOP strategy.