A tight excess risk bound via a unified PAC-Bayesian-Rademacher-Shtarkov-MDL complexity.
Peter D. GrünwaldNishant A. MehtaPublished in: ALT (2019)
Keyphrases
- pac bayesian
- generalization bounds
- worst case
- data dependent
- rademacher complexity
- model selection
- distribution free
- upper bound
- lower bound
- error bounds
- generalization error bounds
- risk bounds
- learning theory
- vc dimension
- generalization ability
- generalization error
- sample size
- function classes
- ranking algorithm
- statistical learning theory
- computational complexity
- machine learning
- linear classifiers
- learning problems
- statistical learning
- uniform convergence
- learning algorithm