mlsanityData to 7 Oct 2026

TFGformer: Multivariate Time Series Forecasting via Time-Frequency Graph Learning and Covariate Fusion

arXiv 2607.29459 ↗First posted 31 Jul 2026cs.LG cs.AIcited by 0

On the map: Retrieval-augmented generation → Time series and agricultural RAG

Abstract (arXiv)

Large-scale multivariate time series from heterogeneous IoT sensors demand accurate long-term forecasting for resource scheduling and predictive maintenance. While recent time series foundation models exhibit strong generalization, they rely on static parametric knowledge and lack dynamic access to external historical patterns during inference. Retrieval-Augmented Generation (RAG) offers a potential remedy, yet its application to time series forecasting is challenged by magnitude variations across heterogeneous sources and the mismatch between historical similarity and future consistency. We propose CrossRAG, a retrieval-augmented forecasting framework that integrates Shape-Aware Memory (SAM) with RevIN normalization for magnitude-robust shape-level retrieval, Future-Consistent Contrastive (FCC) learning to distinguish informative references from hard negatives with similar history but divergent futures, and Cross-Attention Temporal Fusion (CATF) to fuse retrieved historical--future reference pairs into the backbone's representations at the representation level. Experiments on seven public benchmarks show that CrossRAG consistently outperforms both parametric-only baselines and existing retrieval-augmented forecasting methods.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Time series and agricultural RAG, one of 15 directions in Retrieval-augmented generation. Builds on 15 of its 18 references that are in the map. First posted on arXiv 31 Jul 2026.

Builds on

15 of 18 references.

References · 15

  1. 2017Attention Is All You Needcited by 59,932
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  3. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,708
  4. 2020Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecastingcited by 2,141
  5. 2022A Time Series is Worth 64 Words: Long-term Forecasting with Transformerscited by 1,981
  6. 2022Are Transformers Effective for Time Series Forecasting?cited by 1,400
  7. 2022FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecastingcited by 1,200
  8. 2022TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysiscited by 1,030
  9. 2023iTransformer: Inverted Transformers Are Effective for Time Series Forecastingcited by 906
  10. 2023Long-term Forecasting with TiDE: Time-series Dense Encodercited by 327
  11. 2019Recurrent Neural Networks for Time Series Forecasting: Current Status and Future Directionscited by 172
  12. 2023Spatio-Temporal Graph Neural Networks for Predictive Learning in Urban Computing: A Surveycited by 140
3 more
  1. 2024DUET: Dual Clustering Enhanced Multivariate Time Series Forecastingcited by 104
  2. 2024Timer-XL: Long-Context Transformers for Unified Time Series Forecastingcited by 83
  3. 2025ChronosX: Adapting Pretrained Time Series Models with Exogenous Variablescited by 21

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Channel-wise Retrieval for Multivariate Time Series Forecasting
  2. 2026Stationarity-Aware Retrieval-Augmented Time Series Forecasting
  3. 2026Factorize to Generalize: Retrieval-Guided Invariant-Dynamic Decomposition for Time Series Forecasting
  4. 2026Not All Retrievals are Useful: Cross-Attention for Input-Aware RAG in Time Series Forecasting
  5. 2025TS-RAG: Retrieval-Augmented Generation based Time Series Foundation Models are Stronger Zero-Shot Forecaster
  6. 2025QuiZSF: A Retrieval-Augmented Framework for Zero-Shot Time Series Forecasting
  7. 2026Spectral Retrieval-Augmented Time-Series Forecasting
  8. 2026RATL: Learning from Retrieved Residuals for Robust Multivariate Time-Series Forecasting