Diversity of Genes, Its Representation and its Use for Creating New Individuals

The paper presents a metric to express the degree of diversity in a population of genes and proposes a structural representation of a population\´s similarity respective diversity by a pattern set. However, diversity cannot be understood as the presence of various genotypes only, but also as their quantitative distribution. To quantify the degree of distribution, the entropy of the patterns is used. Based on these metrics, three algorithms are introduced here, namely one for generating a new individual, which differs from all other individuals in the population to a required degree (we call it "exotic" here), an algorithm, which creates a new individual, which improves the entropy measure, and an algorithm, which creates a new individual from the fittest pattern.