Label-efficient Time Series Representation Learning: A Review

TL;DR

This survey reviews recent methods for label-efficient time series representation learning, introducing a new taxonomy based on in-domain and cross-domain approaches, and discusses future research directions.

Contribution

It presents the first taxonomy categorizing existing approaches as in-domain or cross-domain, and reviews recent advances, limitations, and future directions in the field.

Findings

In-domain and cross-domain strategies differ in external data reliance.

Recent methods improve representation quality with limited labels.

Identifies key limitations and promising future research areas.

Abstract

Label-efficient time series representation learning, which aims to learn effective representations with limited labeled data, is crucial for deploying deep learning models in real-world applications. To address the scarcity of labeled time series data, various strategies, e.g., transfer learning, self-supervised learning, and semi-supervised learning, have been developed. In this survey, we introduce a novel taxonomy for the first time, categorizing existing approaches as in-domain or cross-domain, based on their reliance on external data sources or not. Furthermore, we present a review of the recent advances in each strategy, conclude the limitations of current methodologies, and suggest future research directions that promise further improvements in the field.