Neutralized Empirical Risk Minimization with Generalization Neutrality Bound.
Kazuto FukuchiJun SakumaPublished in: ECML/PKDD (1) (2014)
Keyphrases
- empirical risk minimization
- generalization bounds
- uniform convergence
- vc dimension
- learning theory
- data dependent
- generalization ability
- statistical learning theory
- risk bounds
- model selection
- ranking algorithm
- upper bound
- large deviations
- learning problems
- linear classifiers
- learning machines
- lower bound
- inductive inference
- learning rate
- real valued
- worst case
- empirical risk
- multi class
- support vector
- sample complexity
- upper and lower bounds
- kernel machines
- theoretical framework
- support vector machine svm
- risk minimization