Author(s):
1. Muhammad Shahbaz Shah:
COMSATS University, Abbottabad, Pakistan
2. Izaz Ullah Khan:
COMSATS University, Abbottabad, Pakistan
Abstract:
This research envisages addressing the case when the convex separable programming problem has multiple objective functions to be optimized. Thus a novel convex multi-objective convex separable programming technique is developed for multi-objective convex separable programming problems. The convexity requirements ensure an approximate global optimal solution for the problem only if it is minimizing a convex function or maximizing a concave function subject to a convex set. Moreover, the accuracy is compromised only by the stiffness of the piecewise linear approximations adopted in the solution process. The main advantage of the convex separable approach is that the problem is solved using linear programming methods without enforcing the adjacency restrictions of the simplex restricted basis method. The proposed technique is then implemented for the selection of optimal portfolios in the capital market structure. The convex separable programming technique is adopted for the minimum variance portfolio optimization problem. After that the risk aversion model is studied using the convex separable programming technique. The results of both are then combined into a multi-objective convex separable programming problem and solved with the help of the Python machine learning software. The results obtained identified portfolios that can return more financial benefits to the investors while investing in the capital market structure. Furthermore, the results of the proposed convex separable programming approach were 22.5% greater than the risk aversion model and 17% greater than the minimum risk model. All the three models were solving by using the Python machine learning software. The successful implementation of the technique and the promising results depicts the importance, credibility and usefulness of the technique for identifying optimal portfolio investments to the investors in the capital market structure.
Page(s):
0-0
DOI:
DOI not available
Published:
Journal: First International Conference on Revamped Scientific Outlook of 21st Century (Abstract Book), Volume: 0, Issue: 0, Year: 2022
Keywords:
Multiobjective Convex Separable Programming
,
Capital Market
,
Portfolio Optimization
,
Separable Programming
,
Convex Optimization
References:
References are not available for this document.
Citations
Citations are not available for this document.