Parameter Setting for Multicore CMA-ES with Large Populations

Abstract

The goal of this paper is to investigate on the overall performance of CMA-ES, when dealing with a large number of coresconsidering the direct mapping between cores and individuals-and to empirically find the best parameter strategies for a parallel machine. By considering the problem of parameter setting, we empirically determine a new strategy for CMA-ES, and we investigate whether Self-CMA-ES (a self-adaptive variant of CMA-ES) could be a viable alternative to CMA-ES when using parallel computers with a coarse-grained distribution of the fitness evaluations. According to a large population size, the resulting new strategy for Self-CMA-ES and CMA-ES, is experimentally validated on BBOB benchmark where it is shown to outperform a CMA-ES with default parameter strategy.

A general framework for anisotropic and wavefront moveout parameter analysis

“…Having said this, metaheuristic algorithms still often need tuning of their hyperparameters to efficiently solve a particular problem given a set of time and computational budget constraints. It is widely understood that the performance of metaheuristic optimisation algorithms, given these constraints are strongly correlated with the values given to their hyperparameters (Belkhir et al 2016). In an attempt to optimise the performance of these algorithms, the hyperparameters are found by tuning.…”

“…Such choices may be perfectly fine for the paper as it may not have been in the scope of the paper to do so. However, it has often been shown that tuning can be tremendously beneficial for some metaheuristics (Belkhir et al 2016, Pedersen 2010a, Tanabe and Fukunaga 2015 particularly when computational cost limitations exist, as do in most real-world applications.…”

“…SMAC has also been utilised in successful parameter tuning for multiple budgets pertaining to the Differential Evolution (Tanabe and Fukunaga 2015) metaheuristic and CMA-ES (Belkhir et al 2016) applied to large sets of objective functions.…”