The Degree of Oneness

This paper investigates the use of diploid structures and the possibility of using diploidy in non-standard, non-binary Genetic Algorithms. Unlike the conventional implementation of diploidy in GAs, which typically involves the use of dominance operators, the diploid schemes presented in this paper avoid the use of dominance altogether, while still being rooted firmly in natural biology. The use of diploid structures allows a population to adapt more quickly in unstable environments, which allows the varying of testcases, whether through evolution to gently increase the difficulty of the problem or through some other selection method to reduce testing time. Two new schemes are introduced, additive diploidy and polygenic inheritance, both of which outperform current methods, and both of which are general enough to be applied to a GA with any number of phenotypes, without any need for the use of dominace operators.