This work proposes a novel practical and general-purpose lossless compression algorithm named Neural Markovian Predictive Compression (NMPC), based on a novel combination of Bayesi...
The increasing number of knowledge-based systems that build on a Bayesian belief network or influence diagram acknowledge the usefulness of these frameworks for addressing complex...
Spam messes up user's inbox, consumes network resources and spread worms and viruses. Spam is flooding of unsolicited, unwanted e mail. Spam in blogs is called blog spam or c...
In recent years several techniques have been proposed for modelling the low-dimensional manifolds, or `subspaces', of natural images. Examples include principal component anal...
The Bayesian committee machine (BCM) is a novel approach to combining estimators which were trained on different data sets. Although the BCM can be applied to the combination of a...