Learning recursive probability trees from probabilistic potentials

作者：Cano Andres; Gomez Olmedo Manuel; Moral Serafin; Perez Ariza Cora B^*; Salmeron Antonio

来源：International Journal of Approximate Reasoning, 2012, 53(9): 1367-1387.

DOI：10.1016/j.ijar.2012.06.026

• Bayesian networks
• Probability trees
• Recursive probability trees

A Recursive Probability Tree (RPT) is a data structure for representing the potentials involved in Probabilistic Graphical Models (PGMs). This structure is developed with the aim of capturing some types of independencies that cannot be represented with previous structures. This capability leads to improvements in memory space and computation time during inference. This paper describes a learning algorithm for building RPTs from probability distributions. The experimental analysis shows the proper behavior of the algorithm: it produces RPTs encoding good approximations of the original probability distributions.

• 出版日期2012-12