Quality-diversity in dissimilarity spaces
Artificial Intelligence
2023-11-30 v3 Neural and Evolutionary Computing
Optimization and Control
Abstract
The theory of magnitude provides a mathematical framework for quantifying and maximizing diversity. We apply this framework to formulate quality-diversity algorithms in generic dissimilarity spaces. In particular, we instantiate and demonstrate a very general version of Go-Explore with promising performance.
Keywords
Cite
@article{arxiv.2211.12337,
title = {Quality-diversity in dissimilarity spaces},
author = {Steve Huntsman},
journal= {arXiv preprint arXiv:2211.12337},
year = {2023}
}
Comments
Added Section 7 (for journal submission to supersede GECCO 2023 at DOI 10.1145/3583131.3590409) which discusses "extremal" diversity at scale zero; some other inconsequential changes