Exploiting label dependencies for improved sample complexity.
Lena ChekinaDan GutfreundAryeh KontorovichLior RokachBracha ShapiraPublished in: Mach. Learn. (2013)
Keyphrases
- sample complexity
- theoretical analysis
- pac learning
- upper bound
- learning algorithm
- learning problems
- special case
- lower bound
- vc dimension
- active learning
- supervised learning
- generalization error
- training examples
- pac learnability
- concept classes
- linear threshold
- sample size
- uniform convergence
- query by committee
- sample complexity bounds
- text categorization
- data sets
- irrelevant features