Abstract
This paper investigates the linkage model building for genetic algorithms. By assuming a given quality of the linkage model, a analytical model of time to convergence is derived. Given the computational cost of building the linkage model, an estimated total computational time is obtained by using the derived time-to-convergence model. The models are empirically verified. The results can be potentially used to decide whether applying a linkage-identification technique is worthwhile and give a guideline to speed up the linkage model building. 1
Users
Please
log in to take part in the discussion (add own reviews or comments).