A PAC-Bayes Sample-compression Approach to Kernel Methods.
Pascal GermainAlexandre LacosteFrançois LavioletteMario MarchandSara ShanianPublished in: ICML (2011)
Keyphrases
- kernel methods
- pac bayes
- support vector
- kernel function
- learning problems
- learning tasks
- feature space
- support vector machine
- machine learning
- risk bounds
- compression ratio
- kernel matrix
- compression algorithm
- compression scheme
- reproducing kernel hilbert space
- multiple kernel learning
- sample size
- high dimensional
- kernel learning
- image processing