A New Model Selection Approach to Hybrid Kernel-Based Estimation.
Anna ScampicchioGianluigi PillonettoPublished in: CDC (2018)
Keyphrases
- model selection
- parameter estimation
- error estimation
- cross validation
- hyperparameters
- selection criterion
- sample size
- regression model
- statistical learning
- mixture model
- meta learning
- generalization error
- statistical inference
- model selection criteria
- bayesian learning
- motion segmentation
- variable selection
- kernel methods
- machine learning
- gaussian process
- bayesian information criterion
- automatic model selection
- parameter determination
- likelihood function
- information criterion
- marginal likelihood
- bayesian model selection
- bayesian methods
- support vector machine
- covariate shift
- image sequences
- learning algorithm