Proceedings of the 5th International Conference on Genetic Algorithms
Applying Self-Organised Criticality to Evolutionary Algorithms
PPSN VI Proceedings of the 6th International Conference on Parallel Problem Solving from Nature
Diversity-Guided Evolutionary Algorithms
PPSN VII Proceedings of the 7th International Conference on Parallel Problem Solving from Nature
Self-adaptive Operator Scheduling Using the Religion-Based EA
PPSN VII Proceedings of the 7th International Conference on Parallel Problem Solving from Nature
Diversity Guided Evolutionary Programming: A novel approach for continuous optimization
Applied Soft Computing
Structured population genetic algorithms: a literature survey
Artificial Intelligence Review
| Hi-index | 0.00 |
Traditionally, the population in Evolutionary Algorithms (EA) is modelled as a simple set of individuals. In recent years, several models have been proposed, which demonstrated that a structured population can improve the performance of EAs. We suggest that religions should be considered as an interesting new source of inspiration in this context, because of their important role in the organisation of human societies. Essential concepts of religions are commandments and religious customs, which influence the behaviour of the individuals. Typically, religious rules include the commitments to reproduce, to believe in no other religions, and to convert non-believers. In this paper, we used these concepts and the struggle between religions as an inspiration for a novel Religion-Based EA (RBEA) with a structured population. In the RBEA, individuals inhabit a two dimensional world in which they can move around and interact with each other. Additional ideas are the religion membership of the individuals, which restricts mating between individuals of different religions and the exchange of individuals between religions by conversion. The result of this design is a simple model with spatial diffusion of genes and self-organised subpopulations with varying population sizes. Our experiments on six numerical benchmark problems showed that the performance of the RBEA was clearly better than the performance of a standard EA and a Diffusion EA.