Visualization tool for a terrain-based genetic algorithm

V.S. Gordon, J. Thein
16th IEEE International Conference on Tools with Artificial Intelligence  
We describe and implement a visualization tool applet for a Terrain-Based Genetic Algorithm (TBGA). The TBGA is a self-tuning version of the Cellular Genetic Algorithm (CGA), wherein various combinations of parameter values appear in different physical locations of the population. The TBGA is useful for solving optimization problems as well as for finding good CGA parameter values. By tallying the number of times a new best individual is found for each location in the population, the applet
more » ... strates the progress of evolution as a gradually evolving terrain map showing effective locations as having increasing altitude. We contrast two methods for using the TBGA to determine good parameter settings. The tool can also help educate users unfamiliar with the TBGA and how it works.
doi:10.1109/ictai.2004.121 dblp:conf/ictai/GordonT04 fatcat:v73mc7epnbdfrbrnclktiu3uqe