Probabilistic inference with noisy-threshold models based on a CP tensor decomposition

0427059 - ÚTIA 2015 RIV US eng J - Journal Article
Vomlel, Jiří - Tichavský, Petr
Probabilistic inference with noisy-threshold models based on a CP tensor decomposition.
International Journal of Approximate Reasoning. Roč. 55, č. 4 (2014), s. 1072-1092. ISSN 0888-613X. E-ISSN 1873-4731
R&D Projects: GA ČR GA13-20012S; GA ČR GA102/09/1278
Institutional support: RVO:67985556
Keywords : Bayesian networks * Probabilistic inference * Candecomp-Parafac tensor decomposition * Symmetric tensor rank
Subject RIV: JD - Computer Applications, Robotics
Impact factor: 2.451, year: 2014
http://library.utia.cas.cz/separaty/2014/MTR/vomlel-0427059.pdf

The specification of conditional probability tables (CPTs) is a difficult task in the construction of probabilistic graphical models. Several types of canonical models have been proposed to ease that difficulty. Noisy-threshold models generalize the two most popular canonical models: the noisy-or and the noisy-and. When using the standard inference techniques the inference complexity is exponential with respect to the number of parents of a variable. More efficient inference techniques can be employed for CPTs that take a special form. CPTs can be viewed as tensors. Tensors can be decomposed into linear combinations of rank-one tensors, where a rank-one tensor is an outer product of vectors. Such decomposition is referred to as Canonical Polyadic (CP) or CANDECOMP-PARAFAC (CP) decomposition. The tensor decomposition offers a compact representation of CPTs which can be efficiently utilized in probabilistic inference.
Permanent Link: http://hdl.handle.net/11104/0233078