The problem of portfolio selection is a very challenging problem in computational finance and has received a lot of attention in last few decades. Selecting an asset and optimal weighting of it from a set of available assets is a critical issue for which the decision maker takes several aspects into consideration. Different constraints like cardinality constraints, minimum buy in thresholds and maximum limit constraint are associated with assets selection. Financial returns associated are often strongly non-Gaussian in character, and exhibit multivariate outliers. Taking these constraints into consideration and with the presence of these outliers we consider a multi-objective problem where the percentage of each available asset is so selected that the total profit of the portfolio is maximized while total risk is minimized. Nondominated Sorting Genetic Algorithm-II is used for solving this multiobjective portfolio selection problem. Performance of the proposed algorithm is carried out by performing different numerical experiments using real-world data.
MISHRA, S. K.; PANDA, G.; MEHER, S.; and MAJHI, R.
"CONSTRAINT ROBUST PORTFOLIO SELECTION BY MULTIOBJECTIVE EVOLUTIONARY GENETIC ALGORITHM,"
International Journal of Electronics Signals and Systems: Vol. 2
, Article 11.
Available at: https://www.interscience.in/ijess/vol2/iss1/11