On the Model Selection for Linear Combination of Step-Type Basis Functions.
Taichi HayasakaKatsuyuki HagiwaraNaohiro TodaShiro UsuiPublished in: ICONIP (1) (1997)
Keyphrases
- model selection
- linear combination
- basis functions
- cross validation
- hyperparameters
- small number
- low rank
- variable selection
- machine learning
- mixture model
- selection criterion
- regression model
- set of basis functions
- marginal likelihood
- generalization error
- gaussian process
- spherical harmonics
- model selection criteria
- information criterion
- state space
- multiscale