Distribution-Independent PAC Learning of Halfspaces with Massart Noise.
Ilias DiakonikolasThemis GouleakisChristos TzamosPublished in: NeurIPS (2019)
Keyphrases
- pac learning
- uniform distribution
- concept classes
- agnostic learning
- membership queries
- target concept
- attribute noise
- decision lists
- classification noise
- computational learning theory
- learning theory
- vc dimension
- exact learning
- concept class
- pac learning model
- learning problems
- sample size
- sample complexity
- pac learnable
- term dnf
- statistical queries
- noise model
- boolean functions
- mistake bound
- dnf formulas
- noisy data
- upper bound
- gaussian distribution
- noise level
- supervised learning
- efficient learning
- lower bound
- feature space
- support vector
- learning algorithm