Selecting the number of components in principal component analysis using cross-validation approximations.
Julie JosseFrançois HussonPublished in: Comput. Stat. Data Anal. (2012)
Keyphrases
- cross validation
- model selection
- principal component analysis
- variable selection
- support vector
- training set
- independent component analysis
- cross validated
- generalization error
- information criterion
- dimensionality reduction
- maximum likelihood
- nearest neighbor
- artificial neural networks
- closed form
- high dimensional
- principal components
- noise level
- feature extraction
- image processing