Generalization bounds for mixing processes via delayed online-to-PAC conversions.
Baptiste AbelesEugenio ClericoGergely NeuPublished in: CoRR (2024)
Keyphrases
- generalization bounds
- vc dimension
- pac bayesian
- data dependent
- learning theory
- generalization ability
- statistical learning theory
- convex combinations
- linear classifiers
- learning problems
- model selection
- upper bound
- sample size
- sample complexity
- uniform convergence
- lower bound
- learning machines
- ranking algorithm
- feature set
- theoretical analysis
- inductive inference
- decision trees
- theoretical framework
- optimal solution