Our aim is to describe a partition of a class by a conjunction of characteristic properties. We use a stepwise top-down binary tree method. At each step we select the best variable and its optimal splitting to optimize simultaneously a discrimination criterion given by a prior partition and a homogeneity criterion. Moreover, this method deals with a data table in which each cell contains a histogram of nominal categories but the method can be extended or reduced to other types of data. The method is illustrated on both simulated data and real data.
Mathieu Vrac hasn't uploaded this paper.
Create a free Academia account to let Mathieu know you want this paper to be uploaded.
