Excluding the best and worst individuals from parent selection

Any crossover operator has both beneficial and detrimental effects: it can bring building blocks together or it can tear them apart. In this paper, we provide evidence that the recombination can be biased towards its more beneficial aspects by modifying the parent selection process. A crude motivation is that the worst individuals do not have valuable building blocks to contribute, and it is too risky to subject the best individuals to crossover and have their building blocks separated. This idea is tested on both the hierarchical if and only if (HIFF) and the one dimensional Ising spin glass (SPIN) fitness functions. Four different parent selection schemes are compared and simulations are shown for both generational and steady state implementations. The results indicate this is a promising direction for future research.