Finite sample guarantees for PCA in non-isotropic and data-dependent noise.
Namrata VaswaniPraneeth NarayanamurthyPublished in: Allerton (2017)
Keyphrases
- data dependent
- finite sample
- generalization bounds
- uniform convergence
- principal component analysis
- sample size
- statistical learning theory
- reproducing kernel hilbert space
- dimensionality reduction
- nearest neighbor
- error bounds
- high dimensional
- feature extraction
- face recognition
- data sets
- feature space
- training data
- learning theory
- generalization ability
- vc dimension
- image processing
- machine learning