On the stability and generalization of neural networks with VC dimension and fuzzy feature encoders.
Linu PintoSasi GopalanP. BalasubramaniamPublished in: J. Frankl. Inst. (2021)
Keyphrases
- vc dimension
- neural network
- generalization bounds
- empirical risk minimization
- fuzzy logic
- learning machines
- uniform convergence
- sample size
- upper bound
- sample complexity
- lower bound
- concept classes
- vapnik chervonenkis dimension
- fuzzy sets
- statistical learning theory
- inductive inference
- concept class
- compression scheme
- feature vectors
- pac learning
- learning theory
- generalization ability
- feature set
- worst case
- data dependent
- model selection
- similarity measure