other projects in the database. Software development data is highly variable, which often result,s in underlying trends being hidden. In order to address this problem, a method of ...
Predictive state representations (PSRs) have recently been proposed as an alternative to partially observable Markov decision processes (POMDPs) for representing the state of a dy...
Matthew Rosencrantz, Geoffrey J. Gordon, Sebastian...
We first formulate multiple targets tracking problem in a dynamic Markov network(DMN)which is derived from a MRFs for joint target state and a binary process for occlusion of dual...
We present two machine learning approaches to information extraction from semi-structured documents that can be used if no annotated training data are available, but there does ex...
The problem of opportunistic access of parallel channels occupied by primary users is considered. Under a continuous-time Markov chain modeling of the channel occupancy by the prim...
Qing Zhao, Stefan Geirhofer, Lang Tong, Brian M. S...