Bounding VC-dimension of neural networks: Progress and prospects.
Marek KarpinskiAngus MacintyrePublished in: EuroCOLT (1995)
Keyphrases
- vc dimension
- upper bound
- neural network
- lower bound
- sample complexity
- covering numbers
- concept classes
- vapnik chervonenkis dimension
- statistical learning theory
- sample size
- worst case
- inductive inference
- distribution free
- generalization bounds
- empirical risk minimization
- pac learning
- concept class
- back propagation
- artificial neural networks
- uniform convergence
- generalization error
- pac learnability
- euclidean space
- neural network model
- special case