EM algorithm and varible Neighborhood Search for fitting Finite Mixture Model parameters.
Adel BessadokPierre HansenAbdelwaheb RebaïPublished in: IMCSIT (2009)
Keyphrases
- em algorithm
- finite mixture model
- expectation maximization
- parameter estimation
- neighborhood search
- maximum likelihood
- maximum likelihood estimation
- mixture model
- gaussian mixture
- probabilistic model
- expectation maximisation
- gaussian mixture model
- density estimation
- simulated annealing
- generative model
- maximum a posteriori
- bayesian framework
- unsupervised learning
- probability density function
- traveling salesman problem
- high dimensional
- variable neighborhood search
- k means
- model based segmentation
- clustering algorithm
- image processing
- information retrieval