Population structure increases the evolvability of genetic algorithms |
| |
Authors: | Felix J. H. Hol Xin Wang Juan E. Keymer |
| |
Affiliation: | 1. Department of Bionanoscience, Kavli Institute of Nanoscience, Delft University of Technology, Delft, The Netherlands;2. Key Laboratory of Systems and Control, Institute of Systems Science, Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing 100190, China |
| |
Abstract: | Populations are shaped by the spatial structure of their environment: space organizes interactions between individuals locally, and gives rise to a global population structure. Both local and global population structures can have a profound influence on the evolutionary dynamics of a population. To characterize this influence, we use genetic algorithms with several distinct contact structures to evolve cellular automata, which perform a density classification task. We find that local contact structures (modeled as graphs with various topologies) that limit the number of breeding partners show greater evolvability than well‐mixed populations. Furthermore, we show that the evolvability of well‐mixed populations is enhanced in a metapopulation setting of coupled subpopulations. © 2012 Wiley Periodicals, Inc. Complexity, 2012 |
| |
Keywords: | population structure cellular automata genetic algorithm evolvability |
|
|