Parameter learning for hybrid Bayesian Networks with Gaussian mixture and Dirac mixture conditional densities.
Peter KrauthausenUwe D. HanebeckPublished in: ACC (2010)
Keyphrases
- parameter learning
- gaussian mixture
- gaussian densities
- bayesian networks
- em algorithm
- gaussian density
- expectation maximization
- probability density function
- probability density
- mixture model
- structure learning
- generative model
- gaussian mixture model
- conditional probabilities
- maximum likelihood
- parameter estimation
- closed form
- probabilistic model
- probability distribution
- statistical learning
- covariance matrix
- mixture of gaussians
- incomplete data
- hidden variables
- graphical models
- conditional random fields
- density estimation
- probabilistic inference
- maximum likelihood estimation
- bayesian framework
- conditional independence
- unsupervised learning
- feature selection
- hidden markov models
- pairwise
- markov random field
- markov networks
- k means
- hyperparameters