Using truncated vine copulas in Supervised Probabilistic Classification

Details

Supervisors

Segers, Johan

Faculty

Faculté des sciences

Degree label

Master [120] en science des données, orientation statistique

Abstract

The objective of this paper is to address the problem of learning effective high-dimensional binary classifiers in domains with continuous features. In this context, we extend the naive Bayes classifier by evaluating the dependence structure between the features given the class with a truncated R-vine copula model.