Combining self-labeling and demand based active learning for non-stationary data streams

TL;DR

This paper introduces a novel online k-NN classifier that combines self-labeling with demand-based active learning to effectively handle scarcely labeled, non-stationary data streams with concept drift.

Contribution

It formalizes the problem of scarcely labeled data streams and proposes a new method integrating self-labeling and demand-based active learning for non-stationary environments.

Findings

Effective handling of drifting data streams with scarce labels

Improved classification accuracy over existing methods

Demonstrated robustness to concept drift

Abstract

Learning from non-stationary data streams is a research direction that gains increasing interest as more data in form of streams becomes available, for example from social media, smartphones, or industrial process monitoring. Most approaches assume that the ground truth of the samples becomes available (possibly with some delay) and perform supervised online learning in the test-then-train scheme. While this assumption might be valid in some scenarios, it does not apply to all settings. In this work, we focus on scarcely labeled data streams and explore the potential of self-labeling in gradually drifting data streams. We formalize this setup and propose a novel online $k$-nn classifier that combines self-labeling and demand-based active learning.