Most belief change operators in the AGM tradition assume an underlying plausibility ordering over the possible worlds which is transitive and complete. A unifying structure for the...
Richard Booth, Thomas Meyer, Chattrakul Sombatthee...
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...
We address document image classification by visual appearance. An image is represented by a variable-length list of visually salient features. A hierarchical Bayesian network is ...
This paper introduces a novel solver, namely cross entropy (CE), into the MRF theory for medical image segmentation. The solver, which is based on the theory of rare event simulati...