Cluster Variational Approximations for Structure Learning of Continuous-Time Bayesian Networks from Incomplete Data.
Dominik LinznerHeinz KoepplPublished in: CoRR (2018)
Keyphrases
- incomplete data
- structure learning
- variational approximation
- bayesian networks
- exact inference
- approximate inference
- em algorithm
- graphical models
- parameter estimation
- parameter learning
- maximum likelihood estimation
- probabilistic graphical models
- variational methods
- probabilistic inference
- conditional independence
- dynamic bayesian networks
- expectation maximization
- probabilistic model
- variational inference
- latent variables
- clustering algorithm
- posterior probability
- maximum likelihood
- belief propagation
- free energy
- posterior distribution
- random variables
- conditional probabilities
- probability distribution
- hyperparameters
- generative model
- markov networks
- bayesian inference
- mixture model
- missing values
- missing data
- data points
- probability density function
- message passing
- maximum a posteriori
- network structure
- transfer learning
- sample size
- knn
- training data