Adaptive genetic algorithm based on population entropy estimating
Rui Jiang, Yupin Luo, Dongcheng Hu, Kwok Yip Szeto · Journal of Tsinghua University(Science and Technology) · 2002
This paper presents a parameter adaptive genetic algorithm based on the entropy estimating to balance exploration and exploitation in the problem's solution space while doing optimization. In the algorithm, the new population in each generation consists of three sub populations: a preserved part, a reproduced part and a randomized part with corresponding parameters introduced to control the size of each part. The parameters can be adjusted adaptively by incorporating population entropy into the algorithm to provide a quantitative measure of the diversity of individuals in the population and by adopting a simple yet practical method to estimate the entropy of a given population. Experimental data show that the algorithm can effectively balance the exploration and exploitation and provides excellent performance with complex problems.