On the Learnability of Rich Function Classes.
Joel RatsabyVitaly MaiorovPublished in: J. Comput. Syst. Sci. (1999)
Keyphrases
- function classes
- agnostic learning
- vc dimension
- real valued functions
- uniform convergence
- data dependent
- learning algorithm
- generalization bounds
- sample complexity
- uniform distribution
- learning problems
- target function
- concept class
- inductive inference
- special case
- pac learning
- linear functions
- pac model
- boolean functions
- reproducing kernel hilbert space
- statistical learning theory
- lower bound
- upper bound
- noise tolerant
- decision lists
- machine learning
- concept classes
- boosting algorithms
- pattern languages
- learning machines
- supervised learning
- worst case