The VC-Dimension versus the Statistical Capacity of Multilayer Networks.
Chuanyi JiDemetri PsaltisPublished in: NIPS (1991)
Keyphrases
- vc dimension
- vapnik chervonenkis dimension
- upper bound
- concept classes
- statistical learning theory
- sample complexity
- lower bound
- covering numbers
- inductive inference
- distribution free
- uniform convergence
- sample size
- empirical risk minimization
- pac learnability
- concept class
- pac learning
- compression scheme
- generalization bounds
- learning process
- data sets
- euclidean space
- similarity search
- worst case
- machine learning