A Deep Dive into Effects of Structural Bias on CMA-ES Performance along Affine Trajectories

TL;DR

This paper investigates how structural biases in the modCMA algorithm influence its performance on various affine landscapes, using extensive configuration testing and advanced analysis tools.

Contribution

It provides a comprehensive analysis of modular components' impact on structural bias and performance, revealing key modules affecting optimization success.

Findings

Identified modules significantly affecting structural bias.

Demonstrated interplay between bias and landscape features.

Used extensive configuration testing and advanced analysis tools.

Abstract

To guide the design of better iterative optimisation heuristics, it is imperative to understand how inherent structural biases within algorithm components affect the performance on a wide variety of search landscapes. This study explores the impact of structural bias in the modular Covariance Matrix Adaptation Evolution Strategy (modCMA), focusing on the roles of various modulars within the algorithm. Through an extensive investigation involving 435,456 configurations of modCMA, we identified key modules that significantly influence structural bias of various classes. Our analysis utilized the Deep-BIAS toolbox for structural bias detection and classification, complemented by SHAP analysis for quantifying module contributions. The performance of these configurations was tested on a sequence of affine-recombined functions, maintaining fixed optimum locations while gradually varying the landscape features. Our results demonstrate an interplay between module-induced structural bias and algorithm performance across different landscape characteristics.