A Transformer-based Framework for Multivariate Time Series Representation Learning

TL;DR

This paper introduces a novel transformer-based framework for unsupervised learning of multivariate time series, achieving state-of-the-art results in regression and classification tasks with limited data and demonstrating the benefits of pre-training.

Contribution

It presents the first transformer-based unsupervised framework for multivariate time series representation learning, surpassing existing methods in accuracy and efficiency.

Findings

Outperforms current unsupervised methods on benchmark datasets.

Exceeds supervised methods in limited data scenarios.

Pre-training provides significant performance gains without extra data.

Abstract

In this work we propose for the first time a transformer-based framework for unsupervised representation learning of multivariate time series. Pre-trained models can be potentially used for downstream tasks such as regression and classification, forecasting and missing value imputation. By evaluating our models on several benchmark datasets for multivariate time series regression and classification, we show that not only does our modeling approach represent the most successful method employing unsupervised learning of multivariate time series presented to date, but also that it exceeds the current state-of-the-art performance of supervised methods; it does so even when the number of training samples is very limited, while offering computational efficiency. Finally, we demonstrate that unsupervised pre-training of our transformer models offers a substantial performance benefit over fully supervised learning, even without leveraging additional unlabeled data, i.e., by reusing the same data samples through the unsupervised objective.