Model selection for neural networks: comparing MDL and NIC.
Guido M. te BrakeJoost N. KokPaul M. B. VitányiPublished in: ESANN (1994)
Keyphrases
- model selection
- neural network
- cross validation
- parameter estimation
- hyperparameters
- sample size
- mixture model
- variable selection
- statistical learning
- regression model
- bayesian learning
- meta learning
- unsupervised learning
- gaussian process
- feature selection
- statistical inference
- information criterion
- motion segmentation
- model selection criteria
- parameter determination
- machine learning
- leave one out cross validation
- bayesian information criterion
- error estimation
- generalization error
- back propagation
- marginal likelihood
- automatic model selection
- selection criterion
- probability distribution