Distributed Anytime Generic Inference

Distributed Anytime Generic Inference

My thesis explores the area of approximate inference using the framework of generic inference. This allows application to a wide range of instances, such as probabilistic graphical models (Bayesian and Markov networks), belief functions, which are used in data fusion from sensor networks and propositional logic, among others. Developed and implemented an anytime inference algorithm for obtaining approximate solutions to the inference problem which improve in quality under greater time allocation.