Self-Supervised Learning for Time Series: Contrastive or Generative?

TL;DR

This paper provides a comprehensive comparison between contrastive and generative self-supervised learning methods for time series, analyzing their frameworks, performance, and practical implications to guide future research and application.

Contribution

It offers a systematic comparison of contrastive and generative SSL methods for time series, including implementation, analysis, and practical recommendations.

Findings

Contrastive methods excel in capturing discriminative features.

Generative methods are better at modeling data distribution.

Insights guide the selection of SSL approaches for time series tasks.

Abstract

Self-supervised learning (SSL) has recently emerged as a powerful approach to learning representations from large-scale unlabeled data, showing promising results in time series analysis. The self-supervised representation learning can be categorized into two mainstream: contrastive and generative. In this paper, we will present a comprehensive comparative study between contrastive and generative methods in time series. We first introduce the basic frameworks for contrastive and generative SSL, respectively, and discuss how to obtain the supervision signal that guides the model optimization. We then implement classical algorithms (SimCLR vs. MAE) for each type and conduct a comparative analysis in fair settings. Our results provide insights into the strengths and weaknesses of each approach and offer practical recommendations for choosing suitable SSL methods. We also discuss the implications of our findings for the broader field of representation learning and propose future research directions. All the code and data are released at \url{https://github.com/DL4mHealth/SSL_Comparison}.