A note on the prior parameter choice in finite mixture models of distributions from exponential families.
María Jesús RufoJacinto MartínC. J. PérezPublished in: Comput. Stat. (2010)
Keyphrases
- exponential family
- mixture model
- density estimation
- mixture modeling
- maximum likelihood
- log likelihood
- em algorithm
- closed form
- statistical models
- missing values
- graphical models
- prior knowledge
- maximum a posteriori
- variational methods
- gaussian mixture model
- order statistics
- generative model
- dirichlet distribution
- probability density function
- parameter estimation
- expectation maximization
- unsupervised learning
- markov chain monte carlo
- hidden variables
- model selection
- probability distribution
- image processing
- model based clustering
- bayesian networks