Novel Ways to Use Search History in Evolutionary Algorithms

Evolutionary algorithm has shown great promise in finding good quality solutions for complex engineering and scientific problems. Recently, we have reported a new type of evolutionary algorithm that makes use of the entire search history to make intelligent, informed decisions to guide the search to the optimal solution. Substantial improved search performance has been attained. This research shall attempt to extend the applicability of this class of algorithms in the following areas: (1) extension of the algorithm to problems consisting of parameters with unknown range; (2) extension to problems with time varying landscapes. The result of this research shall make the algorithm more generally applicable to practical optimization scenarios.