Fast log-likelihood computation for mixture densities in a high-dimensional feature space.
Peter BeyerleinPublished in: ICSLP (1994)
Keyphrases
- log likelihood
- high dimensional feature space
- em algorithm
- density estimation
- kernel function
- mixture model
- information theoretic
- feature space
- input space
- maximum likelihood
- expectation maximization
- high dimensional
- kernel methods
- feature vectors
- probability density function
- scoring function
- dimensionality reduction
- closed form
- image processing
- decision trees
- machine learning
- low dimensional
- probability distribution
- gaussian mixture model
- data points
- scale space
- support vector
- hyperplane
- graphical models
- support vectors
- mutual information
- clustering algorithm
- principal component analysis