Stationarity-Aware Retrieval-Augmented Time Series Forecasting

• URL: http://arxiv.org/abs/2606.04135v1
• Date: Tue, 02 Jun 2026 18:47:16 GMT
• Title: Stationarity-Aware Retrieval-Augmented Time Series Forecasting
• Abstract summary: Time series forecasting relies on historical patterns.<n>Real-world series often exhibit non-stationarity and regime shifts.<n>We propose Stationarity-Aware Retrieval-Augmented Time Series Forecasting.
• Score: 5.059121667323038
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: 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.

Related papers

• Model-agnostic Retrieval-Augmented Extended Forecasting for time series [8.126818349121537]
  Time series forecasting with pretrained foundation models has demonstrated strong zero-shot capabilities.<n> achieving optimal performance on time series with short or negligible historical data in domain-specific applications typically requires adaptation via either fine-tuning or RAG.<n>This work explores RAG as a more efficient alternative, in particular RAF, and introduces RAEF, a model-agnostic method built upon RAF.
  arXiv  Detail & Related papers  (2026-08-14T07:59:17Z)
• TFGformer: Multivariate Time Series Forecasting via Time-Frequency Graph Learning and Covariate Fusion [12.980172248250257]
  We propose a retrieval-augmented forecasting framework that integrates Shape-Aware Memory (SAM) with RevIN normalization for magnitude-robust shape-level retrieval.<n>Experiments on seven public benchmarks show that CrossRAG consistently outperforms both parametric-only baselines and existing retrieval-augmented forecasting methods.
  arXiv  Detail & Related papers  (2026-07-31T14:24:26Z)
• ALER-TI: Aligned Latent Embedding Retrieval for Time Series Imputation [2.185694185279913]
  ALER-TI is a retrieval-augmented framework that explicitly leverages historical patterns to supplement degraded local context for more reliable missing-value reconstruction.<n>ALER-TI consistently improves strong baseline models and enhances robustness across diverse imputation settings.
  arXiv  Detail & Related papers  (2026-07-08T16:59:38Z)
• Spectral Retrieval-Augmented Time-Series Forecasting [7.885089945243776]
  Time series forecasting leverages historical patterns to predict future values, but traditional methods face challenges when dealing with complex, non-stationary patterns.<n>Retrieval-augmented approaches have emerged as promising solutions by retrieving similar historical patterns to enhance predictions.<n>We propose SpecReTF, a novel retrieval method that converts time series into windowed frequency representations, measuring similarity with a combined metric.
  arXiv  Detail & Related papers  (2026-06-17T15:42:19Z)
• Semantics-Enhanced Retrieval-Augmented Time Series Forecasting [5.059121667323038]
  Time series forecasting models often benefit from historical patterns.<n>Recent research explored retrieving relevant historical time series segments to enhance forecasting.<n>We propose a multimodal approach: a textbfSemantics-textbfEnhanced textbfRetrieval-textbfAugmented Time Series textbfForecasting framework, SERAF.
  arXiv  Detail & Related papers  (2026-06-12T20:32:10Z)
• Factorize to Generalize: Retrieval-Guided Invariant-Dynamic Decomposition for Time Series Forecasting [64.2640656619583]
  Time series foundation models (TSFMs) have recently achieved strong zero-shot forecasting performance through large-scale pretraining and retrieval-augmented prediction.<n>We propose a Retrieval-guided Invariant-Dynamic DEcomposition framework for time series forecasting.
  arXiv  Detail & Related papers  (2026-05-24T07:31:55Z)
• TimeAPN: Adaptive Amplitude-Phase Non-Stationarity Normalization for Time Series Forecasting [67.91113180885601]
  TimeAPN explicitly models and predicts non-stationary factors from both the time and frequency domains.<n>TimeAPN consistently improves long-term forecasting accuracy across multiple prediction horizons.<n>It outperforms state-of-the-art reversible normalization methods.
  arXiv  Detail & Related papers  (2026-03-18T07:21:24Z)
• Seeking SOTA: Time-Series Forecasting Must Adopt Taxonomy-Specific Evaluation to Dispel Illusory Gains [17.650604674194984]
  We argue that the current practice of evaluating AI/ML time-series forecasting models obscures real progress by overlooking the performance of efficient classical methods.<n>We demonstrate that these "standard" datasets often exhibit dominant autocorrelation patterns and seasonal cycles that can be effectively captured by simpler linear or statistical models.
  arXiv  Detail & Related papers  (2026-03-16T16:34:26Z)
• MEMTS: Internalizing Domain Knowledge via Parameterized Memory for Retrieval-Free Domain Adaptation of Time Series Foundation Models [51.506429027626005]
  Memory for Time Series (MEMTS) is a lightweight and plug-and-play method for retrieval-free domain adaptation in time series forecasting.<n>Key component of MEMTS is a Knowledge Persistence Module (KPM), which internalizes domain-specific temporal dynamics.<n>This paradigm shift enables MEMTS to achieve accurate domain adaptation with constant-time inference and near-zero latency.
  arXiv  Detail & Related papers  (2026-02-14T14:00:06Z)
• Optimal Look-back Horizon for Time Series Forecasting in Federated Learning [26.070107882914844]
  This paper presents a principled framework for adaptive horizon selection in federated time series forecasting.<n>We derive a decomposition of the forecasting loss into a Bayesian term, which reflects irreducible uncertainty.<n>We prove that the total forecasting loss is minimized at the smallest horizon where the irreducible loss starts to saturate, while the approximation loss continues to rise.
  arXiv  Detail & Related papers  (2025-11-16T21:46:54Z)
• A Unified Frequency Domain Decomposition Framework for Interpretable and Robust Time Series Forecasting [81.73338008264115]
  Current approaches for time series forecasting, whether in the time or frequency domain, predominantly use deep learning models based on linear layers or transformers.<n>We propose FIRE, a unified frequency domain decomposition framework that provides a mathematical abstraction for diverse types of time series.<n>Fire consistently outperforms state-of-the-art models on long-term forecasting benchmarks.
  arXiv  Detail & Related papers  (2025-10-11T09:59:25Z)
• MFRS: A Multi-Frequency Reference Series Approach to Scalable and Accurate Time-Series Forecasting [51.94256702463408]
  Time series predictability is derived from periodic characteristics at different frequencies.<n>We propose a novel time series forecasting method based on multi-frequency reference series correlation analysis.<n> Experiments on major open and synthetic datasets show state-of-the-art performance.
  arXiv  Detail & Related papers  (2025-03-11T11:40:14Z)
• Retrieval Augmented Time Series Forecasting [19.71444850959715]
  We argue that the dynamic and event-driven nature of time-series data makes RAG a crucial component of time-series forecasting. We introduce a principled RAG framework for time-series forecasting, called Retrieval Augmented Forecasting (RAF)
  arXiv  Detail & Related papers  (2024-11-12T23:55:11Z)
• Timer-XL: Long-Context Transformers for Unified Time Series Forecasting [67.83502953961505]
  We present Timer-XL, a causal Transformer for unified time series forecasting.<n>Based on large-scale pre-training, Timer-XL achieves state-of-the-art zero-shot performance.
  arXiv  Detail & Related papers  (2024-10-07T07:27:39Z)
• TimeSiam: A Pre-Training Framework for Siamese Time-Series Modeling [67.02157180089573]
  Time series pre-training has recently garnered wide attention for its potential to reduce labeling expenses and benefit various downstream tasks. This paper proposes TimeSiam as a simple but effective self-supervised pre-training framework for Time series based on Siamese networks.
  arXiv  Detail & Related papers  (2024-02-04T13:10:51Z)
• Towards Diverse and Coherent Augmentation for Time-Series Forecasting [22.213927377926804]
  Time-series data augmentations mitigate the issue of insufficient training data for deep learning models. We propose to combine Spectral and Time Augmentation for generating more diverse and coherent samples. Experiments on five real-world time-series datasets demonstrate that STAug outperforms the base models without data augmentation.
  arXiv  Detail & Related papers  (2023-03-24T19:40:34Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.