We study the scalability and efficiency of a GA that we developed earlier to solve the practical cartographic problem of labeling a map with point features. We argue that the special characteristics of our GA make that it fits in well with theoretical models predicting the optimal population size (the Gambler’s Ruin model) and the number of generations until convergence. We then verify these predictions experimentally. It turns out that our algorithm indeed performs according to the theory, leading to a scale-up for the total amount of computational effort that is linear in the problem size.
CITATION STYLE
Van Dijk, S., Thierens, D., & De Berg, M. (2000). Scalability and efficiency of genetic algorithms for geometrical applications. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1917, pp. 683–692). Springer Verlag. https://doi.org/10.1007/3-540-45356-3_67
Mendeley helps you to discover research relevant for your work.