Portfolio selection using ant colony algorithm and entropy optimization

The main objective of portfolio management was to select a suitable stock portfolio. in this study,we used a novel model based on ant colony algorithm and entropy optimization to select an optimal portfolio. meanwhile we introduced a new definition of covariance to consider a difference between the kinds of influences that the shares have on each other. we first reviewed the literature as well as different studies on the subject,and found the four the most important criteria to be more suited for our research objectives. of all companies listed at tehran stock exchange,data of 55 companies from 2003 to 2013 were analyzed as the statistical sample; these companies had the highest return on assets and the lowest risk level. ultimately,we predicted several portfolios containing the best companies with respect to the mentioned criteria. the matlab software application was then used for data analysis,and after developing the entropy decision matrix,portfolios with the lowest entropy were selected. © 2017 pakistan journal of statistics.