With the index weight matrix W, features fuzzy matrix R and S, fuzzy identification matrix U, definition: )(),( hjjhjhjsrwusrd − = (6) As sample j and pattern of the generalized right between h weighted distance. This distance completely describe the sample j and pattern h differences, because it not only considers the index weight matrix, and at the same time to sample j affiliated to the mode of membership as weights h, complete with in fuzzy mathematics can be defined in membership for weight point of view. In order to solve the optimal fuzzy recognition matrix U, establish the objective function:
đang được dịch, vui lòng đợi..
