TS2Vec: Towards Universal Representation of Time Series
This work addresses the challenge of robust time series analysis for domains like forecasting and anomaly detection, offering a novel method that outperforms previous approaches.
The paper tackles the problem of learning universal time series representations by introducing TS2Vec, a hierarchical contrastive learning framework that achieves significant improvements over existing SOTA methods on 125 UCR and 29 UEA datasets for classification, and also excels in forecasting and anomaly detection tasks.
This paper presents TS2Vec, a universal framework for learning representations of time series in an arbitrary semantic level. Unlike existing methods, TS2Vec performs contrastive learning in a hierarchical way over augmented context views, which enables a robust contextual representation for each timestamp. Furthermore, to obtain the representation of an arbitrary sub-sequence in the time series, we can apply a simple aggregation over the representations of corresponding timestamps. We conduct extensive experiments on time series classification tasks to evaluate the quality of time series representations. As a result, TS2Vec achieves significant improvement over existing SOTAs of unsupervised time series representation on 125 UCR datasets and 29 UEA datasets. The learned timestamp-level representations also achieve superior results in time series forecasting and anomaly detection tasks. A linear regression trained on top of the learned representations outperforms previous SOTAs of time series forecasting. Furthermore, we present a simple way to apply the learned representations for unsupervised anomaly detection, which establishes SOTA results in the literature. The source code is publicly available at https://github.com/yuezhihan/ts2vec.