PORTFOLIO OPTIMIZATION USING PCA CLASSIFICATION AND GENETICS ALGORITHM

Main Article Content

NACEUR RAHMANI
NACEUR KHELIL

Abstract

Portfolio optimization is one of the main investors in financial markets, we present in this paper a new approach to obtain an optimal portfolio, which minimizes the risk for a required profit or maximizing the profit of a given risk. To solve the problem we first introduce a concept of mean absolute deviation risk (MAD). The MAD L1risk function can remove most difficulties associated with the Markowitz’s model. We use heuristic evolutionary algorithm to find the optimal portfolio. We have proposed an approach to find a feasible shares portfolio invested in market based on MAD by using PCA (principal component analysis) and genetic algorithm (GA). This approach is organized in two steps: the first one is to use the PCA classification method to classify the actions into classes. In second step we use an algorithm of optimization called MAD-PAG based on genetic algorithm and mean absolute deviation to minimize the risk measured by the MAD and maximize the value of portfolio.

Keywords:
Mean absolute deviation, Portfolio optimizatio, Principal Component Analysis, Genetic algorithm.

Article Details

How to Cite
RAHMANI, N., & KHELIL, N. (2019). PORTFOLIO OPTIMIZATION USING PCA CLASSIFICATION AND GENETICS ALGORITHM. Journal of Global Economics, Management and Business Research, 11(3), 129-141. Retrieved from http://www.ikprress.org/index.php/JGEMBR/article/view/4565
Section
Original Research Article