Asymptotic predictions of the finite-sample risk of the k-nearest-neighbor classifier.
Robert R. SnappSantosh S. VenkateshPublished in: ICPR (2) (1994)
Keyphrases
- finite sample
- sample size
- statistical learning theory
- uniform convergence
- error bounds
- nearest neighbor
- generalization error
- knn
- parzen window
- generalization bounds
- statistical learning
- supervised classification
- sufficient conditions
- model selection
- upper bound
- learning problems
- learning rate
- data dependent
- theoretical framework
- probability distribution