A DEA-based MOEA/D algorithm for portfolio optimization

被引：10
作者：
Zhou, Zhongbao [1 ]
Liu, Xianghui [1 ]
Xiao, Helu [1 ,2 ]
Wu, Shijian [3 ]
Liu, Yueyue [1 ,4 ]
机构：
[1] Hunan Univ, Sch Business Adm, Changsha 410082, Hunan, Peoples R China
[2] Hunan Normal Univ, Sch Business, Changsha 410081, Hunan, Peoples R China
[3] Shandong Univ Sci & Technol, Coll Econ & Management, Qingdao 266590, Shandong, Peoples R China
[4] Quzhou Univ, Coll Entrepreneurship & Innovat, Quzhou 324000, Peoples R China
来源：
CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS | 2019年 / 22卷 / Suppl 6期
基金：
中国国家自然科学基金;
关键词：
Portfolio optimization; Data envelopment analysis; Multi-objective evolutionary algorithm; Cardinality constraints; EFFICIENCY; PREFERENCE; SELECTION;
D O I：
10.1007/s10586-018-2316-7
中图分类号：
TP [自动化技术、计算机技术];
学科分类号：
0812 ;
摘要：
In this paper, we present a multi-objective genetic algorithm DEA-MOEA/D by integrating decomposition method and DEA (Data Envelopment Analysis) approach. The initial solutions are generated by the DEA approach. Difference operators are adopted as the crossover operator of the parent. We adopt the test functions and portfolio optimization problems to compare the performance of DEA-MOEA/D, FDH-MOGA, MOEA/D and NSGA II. The results show that DEA-MOEA/D performs better than other three algorithms, not only for test functions, but for the portfolio optimization.
引用
收藏
页码：14477 / 14486
页数：10
相关论文
共 50 条
12345