Kernel Extraction via Voted Risk Minimization.
Corinna CortesPrasoon GoyalVitaly KuznetsovMehryar MohriPublished in: FE@NIPS (2015)
Keyphrases
- risk minimization
- loss function
- empirical risk
- uniform convergence
- multi category
- reproducing kernel hilbert space
- kernel function
- kernel methods
- line search
- generalization error
- information extraction
- hinge loss
- support vector
- non stationary
- data sets
- linear classifiers
- active learning
- pairwise
- feature space
- e learning