2005 | OriginalPaper | Buchkapitel
A Tree Structured Classifier for Symbolic Class Description
verfasst von : Edwin Diday, M. Mehdi Limam, Suzanne Winsberg
Erschienen in: Data Analysis and Decision Support
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
We have a class of statistical units from a population, for which the data table may contain symbolic data; that is rather than having a single value for an observed variable, an observed value for the aggregated statistical units we treat here may be multivalued. Our aim is to describe a partition of this class of statistical units by further partitioning it, where each class of the partition is described 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; we also aim to insure that the descriptions obtained describe the units in the class to describe and not the rest of the population. We present a real example.