Towards Robust Model Selection Using Estimation and Approximation Error Bounds.
Joel RatsabyRon MeirVitaly MaiorovPublished in: COLT (1996)
Keyphrases
- error bounds
- model selection
- parameter estimation
- error estimation
- cross validation
- theoretical analysis
- selection criterion
- machine learning
- worst case
- regression model
- sample size
- bayesian learning
- marginal likelihood
- hyperparameters
- gaussian process
- information criterion
- generalization error
- feature selection
- statistical learning
- mixture model
- automatic model selection
- motion segmentation
- bayesian model selection
- wavelet synopses
- statistical inference
- likelihood function
- closed form
- bayesian information criterion
- estimation error
- learning algorithm
- polynomial time approximation
- data mining
- parameter determination
- variable selection
- approximation algorithms