On Graph Reconstruction via Empirical Risk Minimization: Fast Learning Rates and Scalability.
Guillaume PapaAurélien BelletStéphan ClémençonPublished in: NIPS (2016)
Keyphrases
- uniform convergence
- empirical risk minimization
- learning rate
- sufficient conditions
- risk minimization
- statistical learning theory
- reproducing kernel hilbert space
- learning algorithm
- convergence rate
- generalization bounds
- generalization error
- vc dimension
- upper and lower bounds
- theoretical analysis
- gaussian kernels
- active learning
- supervised learning
- upper bound
- large deviations
- data dependent
- ranking functions
- kernel function
- worst case