FLAIRS
2009
13 years 2 months ago
2009
We compare two approaches to Bayesian network inference, called variable elimination (VE) and arc reversal (AR). It is established that VE never requires more space than AR, and n...
ATAL
2008
Springer
2008
Springer
Efficient approximate inference in distributed Bayesian networks for MAS-based sensor interpretation
13 years 6 months ago
The multiply sectioned Bayesian network (MSBN) framework is the most studied approach for distributed Bayesian Network inference in an MAS setting. This paper describes a new fram...
ICML
2005
IEEE
14 years 5 months ago
2005
IEEE
Naive Bayes models have been widely used for clustering and classification. However, they are seldom used for general probabilistic learning and inference (i.e., for estimating an...