A general characterization of evolution-based optimization is its blindness towards the optimization landscape. While this can be an asset for a better global search, it inhibits their computational efficiency which becomes highly limiting for very Earge search spaces. In contrast to evolution, a human-like rational search would first try to achieve some general and approximate information about the. search space before the search begins. While this information may be imprecise, human complements it. with approximate reasoning to reduce the size of the search space as much as possible. In this paper, we apply computing with words and perceptions to add this human-like rationality to evolutionary optimization methods to efficiently reduce the search space, and then the evolutionary method begins the search in the reduced space. The proposed method is applied to two sets of problems and compared with conventional GA. The results demonstrate that the proposed approach significantly increases the speed of the evolutionary optimization methods, especially when the search space is very large and complex.

Keywords