Heuristic possibilistic clustering and switching regression

Abstract

The paper deals in a preliminary way with the problem of forming of the switching regression models. The proposed approach is based on the clustering of the switching regression data set using a heuristic D-AFC-TC-algorithm of possibilistic clustering. An idea of the D-AFC-TC-algorithm is presented and the method of switching regression models forming is outlined. A numerical example is given and the result of application of the D-AFC-TC-algorithm to the two-dimensional data set is considered. Preliminary conclusions are stated.