Zero-Shot Time Series Forecasting with Covariates via In-Context Learning

TL;DR

COSMIC is a novel zero-shot time series forecasting model that effectively incorporates covariates through in-context learning, achieving state-of-the-art results without requiring covariate-labeled datasets.

Contribution

The paper introduces COSMIC, the first zero-shot forecasting model to leverage covariates via in-context learning, with a new augmentation method that eliminates the need for covariate datasets.

Findings

COSMIC outperforms existing models in zero-shot forecasting tasks.

Effective covariate utilization improves forecasting accuracy.

The model works well with and without covariates.

Abstract

Pretrained time series models, capable of zero-shot forecasting, have demonstrated significant potential in enhancing both the performance and accessibility of time series forecasting. However, existing pretrained models either do not support covariates or fail to incorporate them effectively. We introduce COSMIC, a zero-shot forecasting model that utilizes covariates via in-context learning. To address the challenge of data scarcity, we propose Informative Covariate Augmentation, which enables the training of COSMIC without requiring any datasets that include covariates. COSMIC achieves state-of-the-art performance in zero-shot forecasting, both with and without covariates. Our quantitative and qualitative analysis demonstrates that COSMIC effectively leverages covariates in zero-shot forecasting.