Data-driven reference model selection and application to L-DDC design.
Pauline KergusMartine OliviCharles Poussot-VassalFabrice DemourantPublished in: CoRR (2019)
Keyphrases
- motion segmentation
- model selection
- data driven
- cross validation
- hyperparameters
- parameter estimation
- selection criterion
- variable selection
- sample size
- statistical learning
- bayesian learning
- meta learning
- regression model
- machine learning
- generalization error
- gaussian process
- statistical inference
- mixture model
- information criterion
- error estimation
- bayesian information criterion
- feature selection
- bayesian methods
- mutual information
- least squares
- image registration
- model selection criteria