Battling the Non-stationarity in Time Series Forecasting via Test-time Adaptation

TL;DR

This paper introduces TAFAS, a test-time adaptation framework for time series forecasting that improves model robustness against non-stationarity and distribution shifts by leveraging partial ground truth and a gated calibration module.

Contribution

The paper presents a novel test-time adaptation method, TAFAS, specifically designed for time series forecasting to handle non-stationarity and distribution shifts during deployment.

Findings

TAFAS improves forecasting accuracy under distribution shifts.

The approach is effective across diverse datasets and architectures.

Long-term forecasting benefits significantly from TAFAS.

Abstract

Deep Neural Networks have spearheaded remarkable advancements in time series forecasting (TSF), one of the major tasks in time series modeling. Nonetheless, the non-stationarity of time series undermines the reliability of pre-trained source time series forecasters in mission-critical deployment settings. In this study, we introduce a pioneering test-time adaptation framework tailored for TSF (TSF-TTA). TAFAS, the proposed approach to TSF-TTA, flexibly adapts source forecasters to continuously shifting test distributions while preserving the core semantic information learned during pre-training. The novel utilization of partially-observed ground truth and gated calibration module enables proactive, robust, and model-agnostic adaptation of source forecasters. Experiments on diverse benchmark datasets and cutting-edge architectures demonstrate the efficacy and generality of TAFAS, especially in long-term forecasting scenarios that suffer from significant distribution shifts. The code is available at https://github.com/kimanki/TAFAS.