Time-Series Contrastive Learning against False Negatives and Class Imbalance

TL;DR

This paper identifies fundamental issues of false negatives and class imbalance in time-series contrastive learning, proposing a graph-based semi-supervised method to improve representation quality and class balance.

Contribution

It introduces a simple, adaptable modification to the SimCLR framework that mitigates false negatives and enhances minority class representation using graph structures and semi-supervised learning.

Findings

Significant performance improvements over existing methods on four datasets.

Effective mitigation of false negatives through instance graph construction.

Enhanced minority class representation via semi-supervised classification.

Abstract

As an exemplary self-supervised approach for representation learning, time-series contrastive learning has exhibited remarkable advancements in contemporary research. While recent contrastive learning strategies have focused on how to construct appropriate positives and negatives, in this study, we conduct theoretical analysis and find they have overlooked the fundamental issues: false negatives and class imbalance inherent in the InfoNCE loss-based framework. Therefore, we introduce a straightforward modification grounded in the SimCLR framework, universally adaptable to models engaged in the instance discrimination task. By constructing instance graphs to facilitate interactive learning among instances, we emulate supervised contrastive learning via the multiple-instances discrimination task, mitigating the harmful impact of false negatives. Moreover, leveraging the graph structure and few-labeled data, we perform semi-supervised consistency classification and enhance the representative ability of minority classes. We compared our method with the most popular time-series contrastive learning methods on four real-world time-series datasets and demonstrated our significant advantages in overall performance.