COLT
2008
Springer
14 years 1 months ago
2008
Springer
We study distribution-dependent, data-dependent, learning in the limit with adversarial disturbance. We consider an optimization-based approach to learning binary classifiers from...