Marc Richards's research while affiliated with Colorado State University and other places

Publication (1)

Full-text available
In contrast with the diverse array of genetic algorithms, the Genetic Programming (GP) paradigm is usually applied in a relatively uniform manner. Heuristics have developed over time as to which replacement strategies and selection methods are best. The question addressed in this paper is relatively simple: since there are so many variants of evolu...


... In this way, a robust individual, once on the slopes of the high narrow peak, will be less likely to produce offspring prone to " falling off " in the event of a small change to its structure. Another investigation by Whitley et al. [23] compared steady-state with generational replacement strategies using tournament sizes of 2 and 7 on several popular GP problems including Artificial Ant, 11 Multiplexer and a symbolic regression problem. The results of that study showed that a generational strategy with tournament size of 2 was the worst performing whereas the steady-state strategy with tournament size of 2 was best overall. ...