Not All are Made Equal: Consistency of Weighted Averaging Estimators Under Active Learning

TL;DR

This paper investigates the consistency of weighted averaging estimators in active learning, proposing a method to restore consistency by incorporating random sampling, and analyzing their behavior under noise conditions.

Contribution

It generalizes Stone's Theorem to active learning, providing conditions for estimator consistency with and without noise, and explains differing behaviors of estimators under these conditions.

Findings

Consistency can be achieved for k-NN, histogram, and kernel estimators under active learning with random sampling.

In noise-free settings, classical consistency results extend to active learning scenarios.

In noisy settings, some estimators remain consistent while others become inconsistent, depending on specific conditions.

Abstract

Active learning seeks to build the best possible model with a budget of labelled data by sequentially selecting the next point to label. However the training set is no longer \textit{iid}, violating the conditions required by existing consistency results. Inspired by the success of Stone's Theorem we aim to regain consistency for weighted averaging estimators under active learning. Based on ideas in \citet{dasgupta2012consistency}, our approach is to enforce a small amount of random sampling by running an augmented version of the underlying active learning algorithm. We generalize Stone's Theorem in the noise free setting, proving consistency for well known classifiers such as $k$-NN, histogram and kernel estimators under conditions which mirror classical results. However in the presence of noise we can no longer deal with these estimators in a unified manner; for some satisfying this condition also guarantees sufficiency in the noisy case, while for others we can achieve near perfect inconsistency while this condition holds. Finally we provide conditions for consistency in the presence of noise, which give insight into why these estimators can behave so differently under the combination of noise and active learning.