Active Learning via Regression Beyond Realizability

Ved Sriraman, Atul Ganju, Shashaank Aiyer, Karthik Sridharan

Abstract:

We present a new active learning framework for multiclass classification based on surrogate risk
minimization that operates beyond the standard realizability assumption. Existing surrogate-based
active learning algorithms crucially rely on realizability—the assumption that the optimal surrogate
predictor lies within the model class—limiting their applicability in practical, misspecified settings.
In this work we show that under conditions significantly weaker than realizability, as long as the
class of models considered is convex, one can still obtain a label and sample complexity comparable
to prior work. Despite achieving similar rates, the algorithmic approaches from prior works can be
shown to fail in non-realizable settings where our assumption is satisfied. Our epoch-based active
learning algorithm departs from prior methods by fitting a model from the full class to the queried
data in each epoch and returning an improper classifier obtained by aggregating these models.