Multivariate-bounded Gaussian mixture model with minimum message length criterion for model selection.
Muhammad AzamNizar BouguilaPublished in: Expert Syst. J. Knowl. Eng. (2021)
Keyphrases
- gaussian mixture model
- model selection
- finite mixtures
- mixture model
- minimum message length
- unsupervised learning
- expectation maximization
- selection criterion
- model selection criteria
- bayesian information criterion
- em algorithm
- cross validation
- regression model
- marginal likelihood
- hyperparameters
- sample size
- feature selection
- parameter estimation
- density estimation
- variable selection
- information criterion
- background subtraction
- mixture components
- gaussian mixture
- generalization error
- machine learning
- gaussian distribution
- generative model
- probabilistic model
- gaussian process
- maximum likelihood
- multivariate regression
- statistical tests
- latent variables
- mixture of gaussians
- supervised learning
- feature vectors
- feature space
- support vector
- information retrieval
- data mining