63 Double-Stage Portfolio Selection Using K-Means Clustering and Genetic Algorithms

In this study, a double-stage process is proposed for portfolio selection. In the first stage, a clustering method is used to identify good quality assets in terms of 1-reliability asset classification. In the second stage, investment allocation in the selected good quality assets is optimized using a genetic algorithm based on stochastic model of portfolio selection with minimum transaction lot and it's the 1-reliability decision. Through the two-stage 1-reliability model process, an optimal portfolio can be determined. The experimental result has showed that its application in portfolio selection is reliable and useful.