Hybrid Multi-Population Genetic Algorithm for Multi-Criteria Portfolio Selection

Mirabi, Mohammad; Ghaneai, Hossein

doi:DOI‎:10.22034/maco.6.1.5

Hybrid Multi-Population Genetic Algorithm for Multi-Criteria Portfolio Selection

Articles in Press

Document Type : Original Article

Authors

Mohammad Mirabi ¹

Hossein Ghaneai ²

¹ Department of Industrial Engineering, Meybod University, Meybod, Iran.

² Department of Computer Engineering, Meybod University, Meybod, Iran.

DOI‎:10.22034/maco.6.1.5

Abstract

One of the most important decision for each investor is how to select the best financial assets to have more profit in future. Portfolio selection is a knowledge-based decision-making process that guide investor to select an optimal or near optimal basket of investment. During recent decades so many approaches are developed in this context. Heuristics and meta-heuristics have the major part of papers that are considered by researchers. Local search, ant colony, simulated annealing and neural network are some of mentioned methods in the literature. In this research, we develop one meta-heuristic method based on genetic and simulates annealing to select portfolio for risky investor with multi criterial approach. The approach that is shown begins with a multipopulation of created solutions and works its way up to the ultimate solution. Studies comparing our approach to other newly developed methods in the literature show that it can locate a strong portfolio with a high rate of return.

Keywords

Portfolio Selection

Genetic Algorithm

Multi Criterial

Genetic Operators

Meta-Heuristic

Subjects