Illinois Genetic Algorithms Laboratory

The development of population-sizing models for genetic algorithms (Goldberg, Deb, & Clark, 1992; Harik, Cantu-Paz, Goldberg, & Miller, 1999) was important for better understanding of genetic algorithms. Last year, Kumara Sastry, Una-May O’Reilly, and David E. Goldberg published a paper on population sizing for genetic programming based upon decision making. I think that this is one of the most important theoretical results in genetic programming and I think that everyone who is seriously interested in understanding genetic programming should have a look at the paper. To let the authors speak for themselves, here’s a part of the abstract of this paper:

This paper derives a population sizing relationship for genetic programming (GP). Following the population-sizing derivation for genetic algorithms in Goldberg, Deb, and Clark (1992), it considers building block decision making as a key facet. The analysis yields a GP-unique relationship because it has to account for bloat and for the fact that GP solutions often use subsolutions multiple times. The population-sizing relationship depends upon tree size, solution complexity, problem di±culty and building block expression probability.