Inference for Generalized Linear Models via alternating directions and Bethe Free Energy minimization.
Sundeep RanganAlyson K. FletcherPhilip SchniterUlugbek S. KamilovPublished in: ISIT (2015)
Keyphrases
- energy minimization
- generalized linear models
- dual decomposition
- loopy belief propagation
- regression model
- graph cuts
- energy function
- belief propagation
- logistic regression
- free energy
- dirichlet process
- markov random field
- message passing
- image segmentation
- active contour model
- curve evolution
- linear regression
- belief networks
- exact inference
- bayesian networks
- mixture model
- graphical models
- probabilistic inference
- gaussian processes
- multi task learning
- approximate inference
- bayesian model
- sampling algorithm
- parameter learning
- bayesian inference
- least squares
- variational inference
- probabilistic model
- pairwise
- level set