Cooperative Coevolution of Automatically Defined Functions with Gene Expression Programming

摘要: The decomposition of problems into smaller elements is a widespread approach. In this paper we consider two approaches that are based over the principle to segmentation for resolution resultant sub-components. On one hand, have Automatically Defined Functions (ADFs), which originally emerged as refinement genetic programming reuse code and modulirize programs components, on other incorporated co evolution implementation ADFs, present cooperative evolutionary-based approach problem developing implemented module Gene Expression Programming (GEP) virtual gene Genetic Algorithm (vgGA) framework, tested ADFs in three symbolic regression problems, comparing it with conventional algorithm. Our results show simple function algorithm performs better than our evolutionary approach, but more complex functions outperformed by Also, an implement GEP minimally invasive way almost any implementation.