Convergence Rates of Smooth Message Passing with Rounding in Entropy-Regularized MAP Inference.
Jonathan N. LeeAldo PacchianoMichael I. JordanPublished in: AISTATS (2020)
Keyphrases
- message passing
- map inference
- convergence rate
- belief propagation
- graphical models
- markov random field
- lp relaxation
- dual decomposition
- approximate inference
- distributed systems
- linear programming
- probabilistic model
- approximation algorithms
- primal dual
- graph cuts
- energy minimization
- least squares
- image segmentation
- image labeling
- markov networks
- pairwise
- maximum a posteriori
- higher order
- belief networks
- cutting plane
- bayesian networks
- objective function
- conditional random fields
- energy function
- parameter estimation
- mixed integer
- semantic segmentation
- upper bound
- learning algorithm