A matrix approach for finding extrema: problems with modularity, hierarchy, and overlap

摘要: Unlike most simple textbook examples, the real world is full with complex systems, and researchers in many different fields are often confronted by problems arising from such systems. Simple heuristics or even enumeration works quite well on small easy problems; however, to efficiently solve large difficult problems, proper decomposition according system key. In this research project, investigating analyzing interactions between components of systems shed some light problem decomposition. By recognizing three bare-bone types interactions---modularity, hierarchy, overlap, theories models developed dissect inspect context genetic algorithms. This dissertation presents a project develop competent optimization method boundedly modularity, overlap explicit The proposed algorithm design utilizes matrix representation an interaction graph analyze decompose problem. results thesis should benefit both technically scientifically. Technically, develops automated dependency structure clustering technique it black-box solver. Scientifically, model better describes helps gain important insights through explicitness procedure.