Now showing items 1-2 of 2
Scalable, Lifted Maximum a Posteriori Inference
This dissertation focuses on Markov logic networks (MLNs), a knowledge representation tool that elegantly unifies first-order logic (FOL) and probabilistic graphical models (PGMs). FOL enables compact representation while ...
Scalable Learning Approaches for Sum-Product-Cutset Networks
Tractable models are a subclass of probabilistic graphical models (PGMs) in which exact inference can be performed tractably – a very desirable property missing from arbitrary PGMs like Bayesian and Markov networks – exact ...