A gradient BYY harmony learning rule on Gaussian mixture with automated model selection.
Jinwen MaTaijun WangLei XuPublished in: Neurocomputing (2004)
Keyphrases
- model selection
- gaussian mixture
- learning rules
- mixture model
- model selection criteria
- selection criterion
- cross validation
- gaussian mixture model
- em algorithm
- neural network
- covariance matrix
- biologically inspired
- artificial neural networks
- sample size
- hyperparameters
- unsupervised learning
- parameter estimation
- automatic model selection
- density estimation
- expectation maximization
- probability density function
- closed form
- marginal likelihood
- feature selection
- back propagation
- machine learning
- gaussian process
- learning scheme
- upper bound
- bayesian information criterion
- image reconstruction
- principal component analysis
- selection criteria
- information criterion
- objective function
- computer vision
- learning algorithm