Newton method for the ICA mixture model.
Jason A. PalmerScott MakeigKenneth Kreutz-DelgadoBhaskar D. RaoPublished in: ICASSP (2008)
Keyphrases
- mixture model
- newton method
- convergence analysis
- regularized least squares
- linear equations
- variational inequalities
- em algorithm
- gaussian mixture model
- optimality conditions
- generative model
- probabilistic model
- model selection
- face recognition
- density estimation
- expectation maximization
- linear svm
- unsupervised learning
- language model
- nonnegative matrix factorization
- feature extraction
- maximum likelihood
- signal processing
- probability density function
- sparse representation
- principal component analysis
- bayesian framework
- logistic regression
- training set