Inheritance-Based Diversity Measures for Explicit Convergence Control in Evolutionary Algorithms

2018/10/30 by Thomas Gabor, Lenz Belzner, Claudia Linnhoff-Popien · 1 citation
Computer Science · #cs.NE

paper · pdf · doi:10.1145/3205455.3205630

GECCO '18: Genetic and Evolutionary Computation Conference, 2018, Kyoto, Japan

arxiv created 2018/10/30 · arxiv updated 2018/10/31

Abstract

Diversity is an important factor in evolutionary algorithms to prevent premature convergence towards a single local optimum. In order to maintain diversity throughout the process of evolution, various means exist in literature. We analyze approaches to diversity that (a) have an explicit and quantifiable influence on fitness at the individual level and (b) require no (or very little) additional domain knowledge such as domain-specific distance functions. We also introduce the concept of genealogical diversity in a broader study. We show that employing these approaches can help evolutionary algorithms for global optimization in many cases.

Cited by

• Surrogate Fitness Metrics for Interpretable Reinforcement Learning