Learning a Classifier when the Labeling Is Known
Learning a Classifier when the Labeling Is Known
S. Ben-David,S. Ben-David
2011 · DOI: 10.1007/978-3-642-24412-4_34
International Conference on Algorithmic Learning Theory · 11 Citations
TLDR
The results imply new sample-size lower bounds for the common agnostic PAC model - a lower bound of Ω(1/e2) on the sample complexity of learning deterministic classifiers, as well as novel results about the utility of unlabeled examples in a semi-supervised learning setup.