AN INTUITIONISTIC FUZZY VERSION OF THE NEAREST PROTOTYPE CLASSIFICATION METHOD, BASED ON A MOVING-OF-PATTERN PROCEDURE

An intuitionistic fuzzy version of one of the statistical nonparametrical methods, the nearest prototype (NT) method, is proposed. The degrees of membership, nonmember-ship and indeterminacy (uncertainty) of a pattern with unknown classification to a given class have been evaluated taking into account the distance to the prototype (mean) vector of that class. The confidence in the classification made is enhanced by means of a procedure of moving of the pattern towards the classes. The moving is based on the correlation coefficients and the distances to the class prototypes. In this way the degrees are being modified and adjusted, getting more precise values and aiming to decrease and possibly eliminate the indeterminacy. The procedure stops when the uncertainty is made negligibly small.