PAC-Bayesian Bound for the Conditional Value at Risk.
Zakaria MhammediBenjamin GuedjRobert C. WilliamsonPublished in: NeurIPS (2020)
Keyphrases
- pac bayesian
- distribution free
- rademacher complexity
- generalization bounds
- data dependent
- error bounds
- normal distribution
- risk bounds
- generalization ability
- vc dimension
- generalization error
- large deviations
- upper bound
- lower bound
- ranking algorithm
- uniform convergence
- learning theory
- bayesian networks
- reproducing kernel hilbert space
- model selection
- objective function