JMLR
2010
12 years 11 months ago
2010
We study the problem of active learning with convex loss functions. We prove that even under bounded noise constraints, the minimax rates for proper active learning are often no b...