一类投资优化模型的遗传算法  

A genetic algorithm for a portfolio optimization model

作　　者：马宇红[1,2] 巩学文[2] 

机构地区：[1]西北师范大学学报编辑部,甘肃兰州730070 [2]西北师范大学数学与信息科学学院,甘肃兰州730070

出　　处：《西北师范大学学报（自然科学版）》2012年第2期19-23,36,共6页Journal of Northwest Normal University(Natural Science)

基　　金：国家自然科学基金资助项目(11061030)

摘　　要：建立了考虑交易费用,并带有整手交易、风险证券投资限额约束和总资本约束的均值-绝对偏差投资优化模型.根据问题可行解的具体特点,提出了一种改进的不可行解的随机修复技巧,并据此设计了一种改进的遗传算法,从而实现了该问题的求解.实证分析表明,本文给出的遗传算法具有较高的搜索效率和稳定性,只要设定适当的遗传代数,从任意初始种群开始,该算法都能够稳定地得到问题的近似最优解.A mean-absolute deviation portfolio optimization model with transaction cost is established,which also take account of several constraints,such as integer transaction,investment limit and total capital.A stochastic repair method of unfeasible solutions is proposed according to the specific characteristics of the solutions,and then a new genetic algorithm based the repair of unfeasible solutions is developed.The empirical analysis shows that the genetic algorithm has higher search efficiency and stability.By setting appropriate genetic generation,starting from arbitrary initial population,the algorithm can obtain an approximate optimal solution of the problem stably.

关 键 词：投资优化 整手交易 交易费用 投资限额 随机修复 遗传算法 

分 类 号：O221.2[理学—运筹学与控制论]