基于自适应并行遗传算法的证券投资组合研究  

Study of portfolio investment based on a self-adaptive parallel genetic algorithm

作　　者：谢鑫[1] 胡云姣[1] 方永峰[1] 

机构地区：[1]北京化工大学理学院,北京100029

出　　处：《北京化工大学学报（自然科学版）》2010年第4期141-144,共4页Journal of Beijing University of Chemical Technology(Natural Science Edition)

摘　　要：将传统的马科维茨模型进行了改进,引入了风险厌恶因子,对投资比例设定了上下限,进一步利用熵对风险进行了修正,并加入了专家评价对模型实行了模糊化处理。同时提出了一种自适应并行遗传算法,其运算时间短,而且随机搜索,其遗传因子能够进行自我调节,不易陷于局部最优。将该算法引入证券投资组合领域,将数据随机分为若干个小组,同时进行遗传优化,提高了运算效率。通过应用实例,求解改进的模型,计算表明自适应并行遗传算法能够准确快速地解决证券投资组合优化问题。An improved portfolio investment approach,based on the traditional Markowitz model,is proposed in this paper,based on the introduction of a risk aversion factor and setting the upper and lower limits of the investment proportion. Moreover,the portfolio risk is amended making use of the entropy criterion,and then the model is fuzzified by incorporating expert evaluation. A self-adaptive parallel genetic algorithm （SPGA） is designed to solve the model. First of all,the data from the portfolio investments are randomly divided into several different groups. Then,the genetic algorithms with self-regulating genetic factors are applied to deal with all the groups simultaneouly. The mechanism of the SPGA has been shown to be a fast random search with a low possibility of getting trapped in the local optima. A numerical experiment was conducted,and the result showed that the SPGA can solve the portfolio problem in an accurate and fast manner.

关 键 词：并发处理 投资组合优化 自我调节 遗传优化 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]