Asymptotic properties of the EM algorithm estimate for normal mixture models with component specific variances.
Dechavudh NityasuddhiDankmar BöhningPublished in: Comput. Stat. Data Anal. (2003)
Keyphrases
- mixture model
- em algorithm
- asymptotic properties
- expectation maximization
- gaussian mixture model
- maximum likelihood
- density estimation
- generative model
- mixture modeling
- maximum likelihood estimation
- parameter estimation
- probabilistic model
- finite mixture models
- likelihood function
- log likelihood
- model based clustering
- mixture components
- language model
- gaussian mixture
- maximum a posteriori
- gaussian distribution
- finite mixtures
- probability density function
- mixture of gaussians
- finite mixture model
- model selection
- probabilistic mixture model
- estimate the model parameters
- machine learning
- unsupervised learning
- posterior probability