Convergent and Correct Message Passing Schemes for Optimization Problems over Graphical Models
Nicholas RuozziSekhar TatikondaPublished in: CoRR (2010)
Keyphrases
- message passing
- graphical models
- belief propagation
- sum product algorithm
- optimization problems
- factor graphs
- approximate inference
- probabilistic inference
- random variables
- evolutionary algorithm
- probabilistic model
- markov networks
- loopy belief propagation
- metaheuristic
- belief networks
- bayesian networks
- probabilistic graphical models
- exact inference
- objective function
- conditional random fields
- structure learning
- conditional independence
- influence diagrams
- knapsack problem
- energy minimization
- distributed systems
- markov random field
- junction tree
- map inference
- free energy
- partition function
- high quality
- computer vision
- branch and bound
- stereo matching
- energy function
- probabilistic reasoning
- max margin
- upper bound
- pairwise