• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

计算机工程与科学

• 人工智能与数据挖掘 • 上一篇    下一篇

基于FDH的分区域多目标遗传算法

周忠宝1,刘悦悦1,2,金倩颖1,肖和录1,程旭曼3   

  1. (1.湖南大学工商管理学院,湖南 长沙 410082;
    2.衢州学院创业学院,浙江 衢州 324000;
    3.长沙南方职业学院经济管理系,湖南 长沙 410208)
  • 收稿日期:2017-03-08 修回日期:2017-05-09 出版日期:2018-07-25 发布日期:2018-07-25
  • 基金资助:

    国家自然科学基金(71771082,71371067);湖南省杰出青年科学基金(2017JJ1012 )

A multi-objective genetic algorithm with
partition strategy based on the FDH model

ZHOU Zhongbao1,LIU Yueyue1,2,JIN Qianying1,XIAO Helu1,CHENG Xuman3   

  1. (1.School of Business Administration,Hunan University,Changsha 410082;
    2.College of Entrepreneurship & Innovation,Quzhou University,Quzhou 324000;
    3.Department of Economics and Management,Changsha Nanfang Professional College,Changsha 410208,China)
  • Received:2017-03-08 Revised:2017-05-09 Online:2018-07-25 Published:2018-07-25

摘要:

提出了一种基于FDH的分区域多目标遗传算法(FDHMOGA)。该算法通过FDH对种群中所有个体进行评价,根据评价所得的效率值和拥挤度对种群进行选择,提高了该算法的局部搜索能力,同时引入分区策略增加算法的搜索范围,有效避免了遗传算法早熟的缺陷,提高了所获解的多样性。对多个测试函数以及投资组合优化问题的测试结果显示,FDHMOGA算法具有良好的计算性能,更具有效性。

关键词: 多目标遗传算法, FDH模型, 分区策略, 投资组合优化

Abstract:

We present a multiobjective genetic algorithm based on the FDH model, called FDHMOGA. The algorithm evaluates the performance of all the individuals of the population and makes choice according to the efficiency value obtained from evaluation and congestion degree, which can improve the local search ability. Meanwhile, we use the partition strategy to enlarge the search range and improve the diversity of the solutions. We adopt several test functions and portfolio optimization models to compare the performance of the FDHMOGA. The results show that the FDHMOGA algorithm has better computation performance and is more effective.
 

Key words: multi-objective genetic algorithm, FDH model, partition strategy, portfolio optimization