Towards Exploratory Quality Diversity Landscape Analysis

TL;DR

This paper investigates the use of Exploratory Landscape Analysis features to characterize Quality Diversity problems, aiming to facilitate automated algorithm selection by understanding how ELA features relate to QD optimization.

Contribution

It is the first study to explore the relationship between ELA features and QD problems, highlighting how various factors influence ELA features in QD contexts.

Findings

ELA features are affected differently in QD compared to random sampling.

Variation operator, behaviour function, archive size, and dimensionality influence ELA features.

Results suggest potential for ELA-based characterization of QD problems.

Abstract

This work is a preliminary study on using Exploratory Landscape Analysis (ELA) for Quality Diversity (QD) problems. We seek to understand whether ELA features can potentially be used to characterise QD problems paving the way for automating QD algorithm selection. Our results demonstrate that ELA features are affected by QD optimisation differently than random sampling, and more specifically, by the choice of variation operator, behaviour function, archive size and problem dimensionality.