Simple Methods for Initializing the EM Algorithm for Gaussian Mixture Models.
Johannes BlömerKathrin BujnaPublished in: CoRR (2013)
Keyphrases
- em algorithm
- gaussian mixture model
- mixture model
- expectation maximization
- density estimation
- maximum likelihood
- gaussian mixture
- finite mixture models
- generative model
- mixture modeling
- probability density function
- incomplete data
- probability density
- likelihood function
- log likelihood
- model based clustering
- parameter estimation
- finite mixtures
- gaussian distribution
- maximum a posteriori
- feature vectors
- maximum likelihood estimation
- hyperparameters
- image segmentation
- mixture distribution
- cross validation