A Progressively Interactive MCDM Method for Portfolio Optimization Problem

Shi-cheng HU, Fang LI, Xing-yu XU, Yang LIU

Abstract


Portfolio is a multi-objective optimization problem with many pareto optimal solutions. In order to assist the DM (Decision Maker) with bounded rationality to get one that is best in his interests, this paper investigated a progressively interactive MCDM (Multi-Criterion Decision Making) method PI-MCDM. The decision tree was adopted to formulate the preference model of the DM. After each interaction, the DM’s feedback was used to update the preference model and then the multi-objective optimization algorithm progresses guided by the updated preference model. The experimental results were compared with those produced by the well-known multi-objective optimization algorithm NSGAII. The comparative results show that the PI-MCDM can give a complete consideration of the DM’s investment preferences and get his most preferred solution.

Keywords


Multi-Objective Optimization, Portfolio Optimization, Preference Model, MCDM


DOI
10.12783/dtcse/aicae2019/31453

Full Text:

PDF

Refbacks

  • There are currently no refbacks.