Portföy Optimizasyonunda Değiştirilmiş Parçacık Sürü Optimizasyonu Yaklaşımı

Abstract

One of the most studied problem in optimization world is portfolio selection problem which is based on knapsack problems. Portfolio optimization is a NP-hard problem. Because of that heuristic methods are preferred solving portfolio selection problem. In this thesis, modified particle swarm is used to solve portfolio optimization which is first application in literature. In this thesis, firstly particle swarm optimization's basic concepts, process and their versions are given. Secondly, Knapsack problems' versions are shown. Thirdly, modified particle swarm optimization and standard particle swarm optimization are used to solve two portfolio model which are cardinality constraint mean-variance model and Sharpe ratio model. In application, standard particle swarm optimization and modified particle swarm optimization are compared by using these models which are cardinality constraint mean-variance model and Sharpe model.