Fuzzy linear regression combining central tendency and possibilistic properties

This paper proposes two fuzzy regression models based on a quadratic programming approach. Fuzzy regression models by linear programming which aim to minimize the sum of the spreads of the estimated intervals prevailed for a decade. Here we propose new fuzzy regression analyses by quadratic programming. In these formulations, the following two objects are considered: minimizing the distances between the estimated output centers and the observed outputs, and minimizing the spreads of the estimated outputs. In order to illustrate our methods, a numerical example is shown in this paper