Learning Recursive Bayesian Multinets for Data Clustering by Means of Constructive Induction

  • Authors:
  • Jose M. Peña;Jose A. Lozano;Pedro Larrañaga

  • Affiliations:
  • Intelligent Systems Group http://www.sc.ehu.es/isg, Department of Computer Science and Artificial Intelligence, University of the Basque Country, P.O. Box 649, E-20080 Donostia-San Seba ...;Intelligent Systems Group http://www.sc.ehu.es/isg, Department of Computer Science and Artificial Intelligence, University of the Basque Country, P.O. Box 649, E-20080 Donostia-San Seba ...;Intelligent Systems Group http://www.sc.ehu.es/isg, Department of Computer Science and Artificial Intelligence, University of the Basque Country, P.O. Box 649, E-20080 Donostia-San Seba ...

  • Venue:
  • Machine Learning - Special issue: Unsupervised learning
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper introduces and evaluates a new class of knowledge model, the recursive Bayesian multinet (RBMN), which encodes the joint probability distribution of a given database. RBMNs extend Bayesian networks (BNs) as well as partitional clustering systems. Briefly, a RBMN is a decision tree with component BNs at the leaves. A RBMN is learnt using a greedy, heuristic approach akin to that used by many supervised decision tree learners, but where BNs are learnt at leaves using constructive induction. A key idea is to treat expected data as real data. This allows us to complete the database and to take advantage of a closed form for the marginal likelihood of the expected complete data that factorizes into separate marginal likelihoods for each family (a node and its parents). Our approach is evaluated on synthetic and real-world databases.