Genetic algorithms (GAs) -- search algorithms based on the mechanics of natural genetics and natural selection -- are playing increasingly important roles in complex systems optimization and machine learning, but further progress depends on the creation of new quantitative and qualitative theory that more adequately explains GA power. This project extends and expands the analytical and metaphorical theory available for genetic algorithm design and analysis. Important topic areas of analytical theory are investigated, including classification methods, deterministic population models, nonconventional population models, and Walsh methods. ANalogies to biological, social, and economic systems are also drawn and used as a source of new GA mechanism. In particular, natural recombination and gene expression are used as a source of design inspiration. Together, the study provides the foundations for understanding the operation of the current generation of genetic algorithms and for designing the next.
