摘要
传统遗传算法(GA)存在着易陷入局部最优的缺陷,本文提出了一种先利用信息熵调整遗传与变异的侧重点,实现算法参数自适应调节,而后再利用小生境算法在基因层面上对GA进行优化以确定最优解的改进的遗传算法。并将此算法引入房地产开发项目投资组合中,计算实例证明了该法具有较高的稳定性和鲁棒性。
An improved GA based on entropy and pheromone is proposed in the paper .The entropy is used to give the scare of the possible optimized zone and to control the selection and crossover. The gene optimized algorithm based on the concept of the pheromone in ACO is used to find the result. The algorithm is introduced into the real estate portfolio, and the experiment results have shown its efficiency in solving the real estate portfolio optimization problems.
出处
《河北建筑科技学院学报》
2005年第1期54-56,共3页
Journal of Hebei Institute of Architectural Science & Technology
基金
河北省教育厅自然科学研究指导计划项目(Z2003404)
