Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Stationarity-Aware Retrieval-Augmented Time Series Forecasting

About

Time series forecasting relies on historical patterns, but real-world series often exhibit non-stationarity and regime shifts that challenge fully parametric forecasters. Inspired by Retrieval-Augmented Generation (RAG), recent work augments forecasters by retrieving relevant historical segments and using them as external evidence at inference time. However, due to the intrinsic non-stationarity of real-world time series, a highly similar past segment does not necessarily imply a similar future, rendering similarity-only retrieval brittle and prone to redundancy. We propose Stationarity-Aware Retrieval-Augmented Time Series Forecasting (SARAF), a framework that adaptively balances relevance and diversity in retrieval. SARAF first forms a candidate pool via temporal similarity with time-aligned enhancement, then applies a diversity-aware selection strategy to cover heterogeneous historical regimes, with the diversification strength automatically modulated by dataset-level stationarity. Moreover, SARAF uses stationarity-aware aggregation to fuse the retrieved futures. Extensive experiments on eight real-world datasets show that SARAF achieves competitive forecasting performance and improves average accuracy and robustness over strong baselines, with particularly clear benefits under challenging non-stationary settings. Code: https://github.com/ShiqiaoZhou/SARAF.

Shiqiao Zhou, Holger Sch\"oner, Zipeng Wu, Edouard Fouch\'e, IAG Wilson, Shuo Wang• 2026

Related benchmarks

TaskDatasetResultRank
Multivariate ForecastingETTh1
MSE0.371
909
Multivariate Time-series ForecastingETTm1
MSE0.346
742
Multivariate Time-series ForecastingETTm2
MSE0.248
593
Multivariate Time-series ForecastingTraffic
MSE0.395
323
Multivariate Time-series ForecastingExchange
MAE0.202
267
Multivariate Time-series ForecastingETTh2
MSE0.34
219
Multivariate Time-series ForecastingElectricity
MAE0.251
109
Multivariate Time-series Forecastingsolar
MAE0.258
104
Showing 8 of 8 rows

Other info

Follow for update