Structural query-by-committee
This work addresses the problem of improving interactive learning efficiency for researchers and practitioners, but it appears incremental as it builds on existing query-by-committee methods.
The authors introduced a framework unifying various interactive learning tasks and generalized the query-by-committee active learning algorithm, analyzing its consistency and convergence rates theoretically and empirically, including in noisy conditions.
In this work, we describe a framework that unifies many different interactive learning tasks. We present a generalization of the {\it query-by-committee} active learning algorithm for this setting, and we study its consistency and rate of convergence, both theoretically and empirically, with and without noise.