Factor Graphs with NUV Priors and Iteratively Reweighted Descent for Sparse Least Squares and More.
Hans-Andrea LoeligerBoxiao MaHampus MalmbergFederico WadehnPublished in: ISTC (2018)
Keyphrases
- least squares
- factor graphs
- sparse linear
- message passing
- graphical models
- belief propagation
- approximate inference
- parameter estimation
- latent variables
- exact inference
- probabilistic graphical models
- probabilistic inference
- prior information
- bayesian framework
- parameter learning
- graph cuts
- optical flow
- bayesian networks
- structure learning
- dynamic bayesian networks
- model construction