3 May 1995, 12:00, WeH 1327
Removing the Genetics from the Standard Genetic Algorithm
Shumeet Baluja
I will present an abstraction of the genetic algorithm (GA), termed
population-based incremental learning (PBIL), that explicitly
maintains the statistics contained in a GA's population, but which
abstracts away the crossover operator and redefines the role of the
population. This results in PBIL being simpler, both computationally
and theoretically, than the GA. Empirical results show that PBIL is
faster and more effective than the GA on a set of 26 commonly used
benchmark problems (PBIL does better than the GAs used for comparison
on 23 out of the 26 problems).
In this talk, however, I will just concentrate on one of the problems
attempted. This problem is custom designed to benefit both from the
GA's crossover operator and from its use of a population. The results
show that PBIL performs as well as, or better than, GAs very carefully
tuned to do well on this problem. This suggests that even on problems
custom designed for GAs, much of the power of the GA may derive from
the statistics maintained implicitly in its population, and not from
the two most distinguishing features of genetic algorithms: the
population and the crossover operator.
Hopefully, there will be quite a bit of time for discussion; questions
and comments are especially welcome.