A Fully Bayesian Inference Approach for Multivariate McDonald's Beta Mixture Model with Feature Selection.
Darya ForouzanfarNarges ManouchehriNizar BouguilaPublished in: CoDIT (2023)
Keyphrases
- mixture model
- fully bayesian
- generative model
- maximum likelihood
- feature selection
- finite mixture models
- dependency parsing
- markov chain monte carlo
- model selection
- em algorithm
- unsupervised learning
- gaussian process
- probabilistic model
- gaussian mixture model
- mixture modeling
- parameter estimation
- density estimation
- regression model
- bayesian framework
- prior knowledge
- expectation maximization
- hyperparameters
- bayesian inference
- posterior distribution
- posterior probability
- probabilistic inference
- feature extraction
- feature space
- probability density function
- belief networks
- bayesian networks
- conditional random fields
- maximum a posteriori
- graphical models
- text mining
- cross validation
- sample size
- machine learning