Tight Bounds for the VC-Dimension of Piecewise Polynomial Networks.
Akito SakuraiPublished in: NIPS (1998)
Keyphrases
- tight bounds
- vc dimension
- upper bound
- piecewise polynomial
- vapnik chervonenkis dimension
- lower bound
- concept classes
- sample complexity
- covering numbers
- distribution free
- statistical learning theory
- worst case
- concept class
- generalization bounds
- sample size
- inductive inference
- empirical risk minimization
- pac learning
- pac learnability
- generalization error
- function classes
- b spline
- image processing