A Transformer-based Framework for Multivariate Time Series Representation Learning
This addresses the problem of learning effective representations from multivariate time series data for researchers and practitioners in fields like forecasting and classification, though it is incremental as it adapts transformers to a new domain.
The authors introduced the first transformer-based framework for unsupervised representation learning of multivariate time series, achieving state-of-the-art performance on benchmark datasets for regression and classification, even with limited training samples, while offering computational efficiency.
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.