An analysis of an embedded crossover scheme on a GA-hard problem

This paper analyzes the effect of an embedded crossover scheme on the performance of a genetic algorithm in solving a GA-hard problem. Embedded crossover schemes have been found to be efficient in achieving convergence in a variety of numerical problems in the past. This analysis confirms these findings through a minimal deceptive problem (with severe deceptive conditions), which would cause a simple GA to diverge from the best solution.