Retrieval Augmented Time Series Forecasting

Kutay Tire, Ege Onur Taga, Muhammed Emrullah Ildiz, Samet Oymak
Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, PMLR 300:4294-4302, 2026.

Abstract

Retrieval-augmented generation (RAG) is a central component of modern LLM systems, particularly in scenarios where up-to-date information is crucial for accurately responding to user queries or when queries exceed the scope of the training data. The advent of time-series foundation models (TSFM), such as Chronos or Moirai, and the need for effective zero-shot forecasting performance across various time-series domains motivates the question: Do the benefits of RAG similarly carry over to time series forecasting? In this paper, we advocate that the dynamic and event-driven nature of time-series data makes RAG a crucial component of TSFMs and introduce a principled RAG framework for time-series forecasting, called Retrieval Augmented Forecasting (RAF). Within RAF, we develop efficient strategies for retrieving related time-series examples and incorporating them into the forecast. Through experiments and mechanistic studies, we demonstrate that RAF improves the forecasting accuracy across diverse time series domains and TSFMs, with gains that are more pronounced for larger models.

Cite this Paper

BibTeX
@InProceedings{pmlr-v300-tire26a, title = { Retrieval Augmented Time Series Forecasting }, author = {Tire, Kutay and Taga, Ege Onur and Ildiz, Muhammed Emrullah and Oymak, Samet}, booktitle = {Proceedings of The 29th International Conference on Artificial Intelligence and Statistics}, pages = {4294--4302}, year = {2026}, editor = {Khan, Emtiyaz and Li, Yingzhen and Solin, Arno and Ramdas, Aaditya}, volume = {300}, series = {Proceedings of Machine Learning Research}, month = {02--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v300/main/assets/tire26a/tire26a.pdf}, url = {https://proceedings.mlr.press/v300/tire26a.html}, abstract = { Retrieval-augmented generation (RAG) is a central component of modern LLM systems, particularly in scenarios where up-to-date information is crucial for accurately responding to user queries or when queries exceed the scope of the training data. The advent of time-series foundation models (TSFM), such as Chronos or Moirai, and the need for effective zero-shot forecasting performance across various time-series domains motivates the question: Do the benefits of RAG similarly carry over to time series forecasting? In this paper, we advocate that the dynamic and event-driven nature of time-series data makes RAG a crucial component of TSFMs and introduce a principled RAG framework for time-series forecasting, called Retrieval Augmented Forecasting (RAF). Within RAF, we develop efficient strategies for retrieving related time-series examples and incorporating them into the forecast. Through experiments and mechanistic studies, we demonstrate that RAF improves the forecasting accuracy across diverse time series domains and TSFMs, with gains that are more pronounced for larger models. } }
Endnote
%0 Conference Paper %T Retrieval Augmented Time Series Forecasting %A Kutay Tire %A Ege Onur Taga %A Muhammed Emrullah Ildiz %A Samet Oymak %B Proceedings of The 29th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2026 %E Emtiyaz Khan %E Yingzhen Li %E Arno Solin %E Aaditya Ramdas %F pmlr-v300-tire26a %I PMLR %P 4294--4302 %U https://proceedings.mlr.press/v300/tire26a.html %V 300 %X Retrieval-augmented generation (RAG) is a central component of modern LLM systems, particularly in scenarios where up-to-date information is crucial for accurately responding to user queries or when queries exceed the scope of the training data. The advent of time-series foundation models (TSFM), such as Chronos or Moirai, and the need for effective zero-shot forecasting performance across various time-series domains motivates the question: Do the benefits of RAG similarly carry over to time series forecasting? In this paper, we advocate that the dynamic and event-driven nature of time-series data makes RAG a crucial component of TSFMs and introduce a principled RAG framework for time-series forecasting, called Retrieval Augmented Forecasting (RAF). Within RAF, we develop efficient strategies for retrieving related time-series examples and incorporating them into the forecast. Through experiments and mechanistic studies, we demonstrate that RAF improves the forecasting accuracy across diverse time series domains and TSFMs, with gains that are more pronounced for larger models.
APA
Tire, K., Taga, E.O., Ildiz, M.E. & Oymak, S.. (2026). Retrieval Augmented Time Series Forecasting . Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 300:4294-4302 Available from https://proceedings.mlr.press/v300/tire26a.html.

Related Material