Improved Multi-label Classification under Temporal Concept Drift: Rethinking Group-Robust Algorithms in a Label-Wise Setting

TL;DR

This paper investigates how group-robust optimization algorithms, originally designed for fairness, can be adapted to handle temporal concept drift in multi-label document classification, showing they outperform traditional resampling methods especially for minority classes.

Contribution

It rethinks group-robust algorithms as adaptation methods for concept drift, demonstrating their effectiveness over sampling-based approaches in multi-label classification with many classes.

Findings

Invariant Risk Minimization outperforms resampling methods.

Spectral Decoupling improves minority class performance.

Effectiveness increases with larger label sets.

Abstract

In document classification for, e.g., legal and biomedical text, we often deal with hundreds of classes, including very infrequent ones, as well as temporal concept drift caused by the influence of real world events, e.g., policy changes, conflicts, or pandemics. Class imbalance and drift can sometimes be mitigated by resampling the training data to simulate (or compensate for) a known target distribution, but what if the target distribution is determined by unknown future events? Instead of simply resampling uniformly to hedge our bets, we focus on the underlying optimization algorithms used to train such document classifiers and evaluate several group-robust optimization algorithms, initially proposed to mitigate group-level disparities. Reframing group-robust algorithms as adaptation algorithms under concept drift, we find that Invariant Risk Minimization and Spectral Decoupling outperform sampling-based approaches to class imbalance and concept drift, and lead to much better performance on minority classes. The effect is more pronounced the larger the label set.