Cluster Variational Approximations for Structure Learning of Continuous-Time Bayesian Networks from Incomplete Data.
Dominik LinznerHeinz KoepplPublished in: NeurIPS (2018)
Keyphrases
- incomplete data
- structure learning
- variational approximation
- bayesian networks
- exact inference
- em algorithm
- approximate inference
- graphical models
- parameter estimation
- parameter learning
- probabilistic inference
- variational methods
- maximum likelihood estimation
- expectation maximization
- conditional independence
- dynamic bayesian networks
- probabilistic model
- probabilistic graphical models
- random variables
- variational inference
- markov networks
- maximum likelihood
- latent variables
- bayesian inference
- hyperparameters
- probability distribution
- posterior probability
- data points
- belief propagation
- image processing
- mixture model
- image segmentation
- clustering algorithm
- belief networks
- missing values
- missing data
- generative model
- free energy
- conditional probabilities
- posterior distribution
- conditional random fields
- input data
- higher order
- knn
- optical flow