Dimension Reduction for Mixtures of Exponential Families.
Shotaro AkahoPublished in: ICANN (1) (2008)
Keyphrases
- dimension reduction
- exponential family
- mixture model
- unsupervised learning
- density estimation
- maximum likelihood
- log likelihood
- high dimensional
- statistical models
- gaussian mixture model
- graphical models
- feature extraction
- missing values
- high dimensional data
- low dimensional
- principal component analysis
- probabilistic model
- em algorithm
- feature selection
- expectation maximization
- generative model
- closed form
- probability density function
- cluster analysis
- feature space
- singular value decomposition
- variational methods
- high dimensionality
- order statistics
- machine learning
- preprocessing
- linear discriminant analysis
- hidden variables
- manifold learning
- model selection
- pattern recognition
- face recognition
- object recognition
- markov chain monte carlo
- nearest neighbor
- markov chain
- bayesian networks