Computing AIC for black-box models using generalized degrees of freedom: A comparison with cross-validation.
Severin HauensteinSimon N. WoodCarsten F. DormannPublished in: Commun. Stat. Simul. Comput. (2018)
Keyphrases
- cross validation
- model selection
- black box
- degrees of freedom
- information criterion
- akaike information criterion
- hyperparameters
- unseen data
- variable selection
- generalization error
- pose estimation
- cross validated
- machine learning
- test cases
- black boxes
- feature selection
- regression problems
- support vector
- end effector
- leave one out cross validation
- mixture model
- meta learning
- selection criterion
- probabilistic model
- learning algorithm
- data sets
- joint space
- nearest neighbor classifiers
- incremental learning