Unsupervized aggregation of commensurate correlated attributes by means of the choquet integral and entropy functionals

In the framework of aggregation by the discrete Choquet integral, the unsupervised method for the identification of the underlying capacity initially put forward in [26] is presented and improvements are proposed. The suggested approach consists in replacing the subjective notion of importance of a subset of attributes by that of information content of a subset of attributes, which can be estimated from the set of profiles by means of an entropy measure. An example of the application of the proposed methodology is given : in the absence of initial preferences, the approach is applied to the evaluation of students.