摘要
论文基于人工蜂群算法的基本思想,对投资组合问题中的带基数约束的均值-方差(CCMV)模型进行求解。在求解过程中,针对CCMV模型设计了能够始终得到可行解的FABC算法。然后对FABC的更新方程进行改进,加入当前最优解的指导作用,得到了收敛速度更快的IFABC算法。最后,基于二次规划对IFABC算法进行进一步改进,提出了求解效果更优的QFABC算法。对真实市场数据进行测试,并对比三种算法的性能,结果表明:QFABC的优化效果最好,但运行时间较长;IFABC算法优化效果接近QFABC算法且运行时间较短。
Based on the idea of artificial bees colony,the mean-variance(CCMV)model with cardinality constraint in portfolio problem is solved in this paper.In the solving process,a FABC algorithm which can always get feasible solution is designed for CCMV model.Then the updated equation of FABC is improved,the guidance of the current optimal solution is added,and a faster convergence rate of IFABC algorithm is gotten.Finally,IFABC algorithm is further improved based on quadratic programming,and a better QFABC algorithm is proposed.Test the market data,and compare the performance of the three algorithms,the results show that QFABC optimization effect is the best,but the running time is long;The optimization effect of IFABC algorithm is close to that of QFABC algorithm and the running time is short.
作者
张懿
ZHANG Yi(Taizhou Vocational and Technical College,Taizhou 318000)
出处
《计算机与数字工程》
2019年第8期1869-1873,1894,共6页
Computer & Digital Engineering
基金
国家自然科学基金项目（编号：71803179）
浙江省教育厅科研项目（编号：Y201533943）
台州市哲学社会科学规划课题成果（编号：18GHQ11）
台州职业技术学院校级青年课题（编号：2016QN06）资助
关键词
人工蜂群算法
投资组合问题
二次规划
artificial bees colony
portfolio problem
quadratic programming
