Correlated Time Series Self-Supervised Representation Learning via Spatiotemporal Bootstrapping
This addresses the challenge of handling large-scale correlated time series in real-world industries like traffic and metro systems, with incremental improvements over existing representation learning models.
The paper tackles the problem of learning efficient representations for correlated time series data to improve forecasting and cold-start transfer, achieving reductions in RMSE, MAE, and MAPE by up to 49% on benchmark datasets.
Correlated time series analysis plays an important role in many real-world industries. Learning an efficient representation of this large-scale data for further downstream tasks is necessary but challenging. In this paper, we propose a time-step-level representation learning framework for individual instances via bootstrapped spatiotemporal representation prediction. We evaluated the effectiveness and flexibility of our representation learning framework on correlated time series forecasting and cold-start transferring the forecasting model to new instances with limited data. A linear regression model trained on top of the learned representations demonstrates our model performs best in most cases. Especially compared to representation learning models, we reduce the RMSE, MAE, and MAPE by 37%, 49%, and 48% on the PeMS-BAY dataset, respectively. Furthermore, in real-world metro passenger flow data, our framework demonstrates the ability to transfer to infer future information of new cold-start instances, with gains of 15%, 19%, and 18%. The source code will be released under the GitHub https://github.com/bonaldli/Spatiotemporal-TS-Representation-Learning