Variational Bayes from the Primitive Initial Point for Gaussian Mixture Estimation.
Yuta IshikawaIchiro TakeuchiRyohei NakanoPublished in: ICONIP (1) (2009)
Keyphrases
- gaussian mixture
- variational bayes
- gaussian mixture model
- hyperparameters
- density estimation
- em algorithm
- expectation maximization
- closed form
- mixture model
- exponential family
- parameter estimation
- latent variables
- log likelihood
- posterior distribution
- probability density function
- bayesian inference
- bayesian learning
- covariance matrix
- maximum likelihood
- bayesian framework
- free energy
- probabilistic model
- model selection
- feature vectors
- feature space
- generative model
- gaussian process
- machine learning
- unsupervised learning