An online classification EM algorithm based on the mixture model.
Allou SaméChristophe AmbroiseGérard GovaertPublished in: Stat. Comput. (2007)
Keyphrases
- mixture model
- em algorithm
- expectation maximization
- density estimation
- gaussian mixture model
- finite mixture models
- maximum likelihood
- generative model
- parameter estimation
- mixture modeling
- model selection
- probabilistic model
- unsupervised learning
- gaussian mixture
- hyperparameters
- feature vectors
- finite mixture model
- maximum likelihood estimation
- likelihood function
- estimate the model parameters
- feature selection
- model based clustering
- decision trees
- feature space
- probability density function
- unlabeled samples
- feature extraction
- log likelihood
- probabilistic mixture model
- text classification
- supervised learning
- incomplete data
- active learning
- pairwise
- training set
- bayesian networks
- image processing
- machine learning