VC-Dimension and Rademacher Averages: From Statistical Learning Theory to Sampling Algorithms.
Matteo RiondatoEli UpfalPublished in: KDD (2015)
Keyphrases
- vc dimension
- statistical learning theory
- generalization bounds
- covering numbers
- data dependent
- risk bounds
- sample size
- empirical risk minimization
- sample complexity
- learning problems
- machine learning
- inductive inference
- worst case
- upper bound
- generalization ability
- lower bound
- uniform convergence
- kernel machines
- concept classes
- statistical learning
- learning theory
- theoretical analysis
- learning algorithm
- model selection
- supervised classification
- learning machines
- theoretical framework
- function classes
- support vector machine
- machine learning algorithms
- dimensionality reduction