In order to analyze uncertain phenomena in real world, the concept of fuzzy random variables is widely employed in model building. In dealing with fuzzy data, defuzzi cation plays a central role. In this paper, portfolio selection problems are dealt as interval values. We calculate the expected values, variance and covariance by using the estimated parameters of underlying probability distribution function. The estimated values enable us to build up a portfolio selection model with estimated parameters on the basic of Markowitz's mean-variance model. The result exempli ed that we have different choices of k which can decide the best expected return and less risk level in our model, also that we can provide not only one choice of portfolio selection but also two or more for decision makers.
International Journal of Innovative Computing, Information and Control, 8(8) , 5935-5944