An Analysis of Phenotypic Diversity in Multi-Solution Optimization

TL;DR

This paper compares different optimization methods in terms of the diversity of solutions they produce, highlighting the strengths of quality diversity and autoencoders in generating highly diverse solution sets.

Contribution

It provides a comparative analysis of solution diversity across multiple optimization techniques and introduces the use of autoencoders to enhance phenotypic diversity.

Findings

Quality diversity produces the most diverse solutions.

Multimodal optimization yields higher fitness solutions.

Autoencoders further increase phenotypic diversity.

Abstract

More and more, optimization methods are used to find diverse solution sets. We compare solution diversity in multi-objective optimization, multimodal optimization, and quality diversity in a simple domain. We show that multiobjective optimization does not always produce much diversity, multimodal optimization produces higher fitness solutions, and quality diversity is not sensitive to genetic neutrality and creates the most diverse set of solutions. An autoencoder is used to discover phenotypic features automatically, producing an even more diverse solution set with quality diversity. Finally, we make recommendations about when to use which approach.