Grid diversity operator for some population-based optimization algorithms.

Salah, Ahmed; Hart, Emma

Abstract

We present a novel diversity method named Grid Diversity
Operator (GDO) that can be incorporated into multiple
population-based optimization algorithms that guides the
containing algorithm in creating new individuals in sparsely
visited areas of the search space. Experimental tests on a set
of unimodal and multimodal benchmark functions from the
literature using GDO in conjunction with opt-aiNet algorithm
show that GDO maintains better diversity in most
cases, leading to an order-of-magnitude reduction in the
number of objective function evaluations needed to converge
while finding similar numbers of peaks in the majority of
benchmarks.

Citation

Salah, A., & Hart, E. (2015). Grid diversity operator for some population-based optimization algorithms. In Proceedings of the Companion Publication of the 2015 on Genetic and Evolutionary Computation Conference - GECCO Companion '15 (1475-1476). https://doi.org/10.1145/2739482.2764664

• Conference Name: GECCO’15 Companion
• Start Date: Jul 11, 2015
• End Date: Jul 15, 2015
• Publication Date: 2015
• Deposit Date: Oct 23, 2015
• Publicly Available Date: May 15, 2017
• Peer Reviewed: Peer Reviewed
• Pages: 1475-1476
• Book Title: Proceedings of the Companion Publication of the 2015 on Genetic and Evolutionary Computation Conference - GECCO Companion '15
• ISBN: 978-1-4503-3488-4
• DOI: https://doi.org/10.1145/2739482.2764664
• Keywords: Artificial Immune Systems; Evolutionary Algorithms; Optimization; Diversity; Grid
• Public URL: http://researchrepository.napier.ac.uk/id/eprint/9210
• Publisher URL: http://dx.doi.org/10.1145/2739482.2764664