Illinois Genetic Algorithms Laboratory

Yu, T.-L., Chen, Y.-P., Goldberg, D. E., Chen, J.-H. (2003)
TR No.: 2003026 | Download PDF | Download PS

Abstract:
This paper proposes an adaptive sampling scheme for genetic algorithms. The adaptive sampling scheme is tested on the sampled OneMax problem. The results suggest that through this scheme, speed-up is obtained for problems with non-uniformly scaled building blocks (BBs). For problems
with uniformly scaled BBs, the proposed adaptive sampling scheme does not give speed-up but still maintains the same performance with respect to the number of function evaluations when the adaptive sampling scheme is not
adopted.