Crossover and diploid dominance with deceptive fitness

A deterministic approach is used to study the effects of recombination and diploid vs. haploid structure on EP efficiency. The deterministic analysis produces an erect result without resorting to multiple trials, at the cost of assuming an infinite population size. An efficiency limit is demonstrated for the recombination (crossover) rate. New evidence of a reduction is shown for the required growth time of diploid vs. haploid. Complete diploid dominance is implemented in a manner which can be applied to any scalar EP, GA or GP problem, stationary or otherwise. In so doing, a dual interpretation of interand intra-gene fitness evaluation becomes apparent, and has a natural extension to vector fitness criteria. Results are consistent with previous nondeterministic, multiple trials that also used stationary fitness criteria