Mixture Functions Based on Deviation and Dissimilarity Functions

Abstract

Mixture functions represent a special class of weighted averaging functions with weights determined by continuous weighting functions which depend on the input values. If they are monotone increasing, they also belong to the important class of aggregation functions. Their construction can be based on minimization of special (weighted) penalty functions using dissimilarity function or based on zero value of the special (weighted) strictly increasing function using deviation functions.