A central problem in learning in complex environmentsis balancing exploration of untested actions against exploitation of actions that are known to be good. The benefit of explora...
Particle filters (PFs) are powerful samplingbased inference/learning algorithms for dynamic Bayesian networks (DBNs). They allow us to treat, in a principled way, any type of prob...
Arnaud Doucet, Nando de Freitas, Kevin P. Murphy, ...
Learning visual models of object categories notoriously requires thousands of training examples; this is due to the diversity and richness of object appearance which requires mode...
Abstract. Suppose that geographic data under investigation are rates associated with polygons. For example, disease incidence, mortality, and census undercount data may be displaye...
In order to implement both the efficiency and security in the Peer-to-Peer (P2P) network, we design a trusted small world overlay P2P network with the role based and reputation bas...