A robust approach based on conditional value-at-risk measure to statistical learning problems.
Akiko TakedaTakafumi KanamoriPublished in: Eur. J. Oper. Res. (2009)
Keyphrases
- learning problems
- learning tasks
- learning algorithm
- supervised learning
- machine learning algorithms
- kernel methods
- computational learning theory
- statistical learning theory
- reinforcement learning
- sample complexity
- semi supervised learning
- information theory
- machine learning
- pac learning
- multi task learning
- multi task
- similarity measure
- data analysis
- multiple kernel learning
- kernel based learning
- classification accuracy
- active learning
- face recognition
- real world