Gaussian mixture parameter estimation with known means and unknown class-dependent variances.
G. R. DattatreyaPublished in: Pattern Recognit. (2002)
Keyphrases
- parameter estimation
- gaussian mixture
- em algorithm
- expectation maximization
- class dependent
- maximum likelihood
- mixture model
- gaussian mixture model
- gaussian distribution
- probability density function
- statistical models
- approximate inference
- model selection
- cost sensitive
- closed form
- least squares
- posterior distribution
- covariance matrices
- generative model
- bayesian framework
- probabilistic model
- unsupervised learning
- markov random field
- image segmentation
- density estimation
- gibbs sampling
- prior knowledge
- data mining
- computer vision
- maximum a posteriori
- covariance matrix
- feature extraction
- bayesian networks