Universal and automatic elbow detection for learning the effective number of components in model selection problems.
Eduardo MorgadoLuca MartinoRoberto San Millán-CastilloPublished in: Digit. Signal Process. (2023)
Keyphrases
- model selection
- statistical learning
- bayesian learning
- unsupervised learning
- cross validation
- meta learning
- hyperparameters
- information criterion
- active learning
- bayesian methods
- covariate shift
- learning algorithm
- learning machines
- error estimation
- variable selection
- parameter estimation
- mixture model
- supervised learning
- automatic model selection
- statistical inference
- model selection criteria
- selection criterion
- gaussian process
- generalization error
- incremental learning
- regression model
- support vector