Output details
11 - Computer Science and Informatics
University of Exeter
Geometric Differential Evolution for Combinatorial and Programs Spaces
<12> This extends a paper that won the best paper award at EuroGP, the premier conference on Genetic Programming. It is published in a top journal, and it is an international collaboration. Vast literature reports ad-hoc generalisation of continuous search algorithms to combinatorial problems and program spaces. This paper introduces for the first time a systematic and rigorous generalisation methodology applied to Differential Evolution to derive specific versions of it for several new representations that perform very well experimentally. The methodology was successfully applied to other algorithms including Particle Swarm Optimisation (Moraglio et al., best paper award nomination at EuroGP).