mlsanityData to 7 Oct 2026

TimeEmb: A Lightweight Static-Dynamic Disentanglement Framework for Time Series Forecasting

arXiv 2510.00461 ↗First posted 1 Oct 2025cs.LG cs.AIcited by 11

Not on the map. Why?

Abstract (arXiv)

Temporal non-stationarity, the phenomenon that time series distributions change over time, poses fundamental challenges to reliable time series forecasting. Intuitively, the complex time series can be decomposed into two factors, ie time-invariant and time-varying components, which indicate static and dynamic patterns, respectively. Nonetheless, existing methods often conflate the time-varying and time-invariant components, and jointly learn the combined long-term patterns and short-term fluctuations, leading to suboptimal performance facing distribution shifts. To address this issue, we initiatively propose a lightweight static-dynamic decomposition framework, TimeEmb, for time series forecasting. TimeEmb innovatively separates time series into two complementary components: (1) time-invariant component, captured by a novel global embedding module that learns persistent representations across time series, and (2) time-varying component, processed by an efficient frequency-domain filtering mechanism inspired by full-spectrum analysis in signal processing. Experiments on real-world datasets demonstrate that TimeEmb outperforms state-of-the-art baselines and requires fewer computational resources. We conduct comprehensive quantitative and qualitative analyses to verify the efficacy of static-dynamic disentanglement. This lightweight framework can also improve existing time-series forecasting methods with simple integration. To ease reproducibility, the code is available at https://github.com/showmeon/TimeEmb.

Key facts. Cited by 11 papers in the map (data through 7 Oct 2026). Builds on 35 of its 57 references that are in the map. First posted on arXiv 1 Oct 2025.

Builds on

35 of 57 references.

References · 35

  1. 2017Attention Is All You Needcited by 59,932
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2020Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecastingcited by 2,141
  4. 2022A Time Series is Worth 64 Words: Long-term Forecasting with Transformerscited by 1,981
  5. 2020Reformer: The Efficient Transformercited by 1,658
  6. 2021Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecastingcited by 1,654
  7. 2022Are Transformers Effective for Time Series Forecasting?cited by 1,400
  8. 2022FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecastingcited by 1,200
  9. 2022TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysiscited by 1,030
  10. 2023iTransformer: Inverted Transformers Are Effective for Time Series Forecastingcited by 906
  11. 2017Modeling Long- and Short-Term Temporal Patterns with Deep Neural Networkscited by 869
  12. 2019N-BEATS: Neural basis expansion analysis for interpretable time series forecastingcited by 744
23 more
  1. 2021SCINet: Time Series Modeling and Forecasting with Sample Convolution and Interactioncited by 335
  2. 2024TimeMixer: Decomposable Multiscale Mixing for Time Series Forecastingcited by 333
  3. 2019Unsupervised Scalable Representation Learning for Multivariate Time Seriescited by 267
  4. 2023TSMixer: An All-MLP Architecture for Time Series Forecastingcited by 232
  5. 2023FITS: Modeling Time Series with 10k Parameterscited by 185
  6. 2023Frequency-domain MLPs are More Effective Learners in Time Series Forecastingcited by 170
  7. 2022Spatial-Temporal Identity: A Simple yet Effective Baseline for Multivariate Time Series Forecastingcited by 149
  8. 2022Decoupled Dynamic Spatial-Temporal Graph Neural Network for Traffic Forecastingcited by 122
  9. 2023Koopa: Learning Non-stationary Time Series Dynamics with Koopman Predictorscited by 110
  10. 2024CALF: Aligning LLMs for Time Series Forecasting via Cross-modal Fine-Tuningcited by 105
  11. 2024CycleNet: Enhancing Time Series Forecasting through Modeling Periodic Patternscited by 102
  12. 2024TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignmentcited by 97
  13. 2024SOFTS: Efficient Multivariate Time Series Forecasting with Series-Core Fusioncited by 75
  14. 2024FilterNet: Harnessing Frequency Filters for Time Series Forecastingcited by 44
  15. 2024Fredformer: Frequency Debiased Transformer for Time Series Forecastingcited by 43
  16. 2024A Unified Replay-based Continuous Learning Framework for Spatio-Temporal Prediction on Streaming Datacited by 39
  17. 2025Time Series Forecasting via Reasoning: A Slow-Thinking Approach with Reinforcement Fine-Tuned LLMscited by 34
  18. 2025Can Slow-thinking LLMs Reason Over Time? Empirical Studies in Time Series Forecastingcited by 19
  19. 2025Towards Cross-Modality Modeling for Time Series Analytics: A Survey in the LLM Eracited by 11
  20. 2015The Fourier Decomposition Method for nonlinear and nonstationary time series analysiscited by 5
  21. 2023T-SaS: Toward Shift-aware Dynamic Adaptation for Streaming Datacited by 3
  22. 2023Real World Time Series Benchmark Datasets with Distribution Shifts: Global Crude Oil Price and Volatilitycited by 1
  23. 2025Temporal Distribution Shift in Real-World Pharmaceutical Data: Implications for Uncertainty Quantification in QSAR Modelscited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 10.

11 papers.

  1. 2026To See Far, Look Close: Evolutionary Forecasting for Long-term Time Seriescited by 1
  2. 2026What If We Let Forecasting Forget? A Sparse Bottleneck for Cross-Variable Dependenciescited by 1
  3. 2026PULSE: Generative Phase Evolution for Non-Stationary Time Series Forecastingcited by 1
  4. 2025Invariant Representation Learning for Source-Free Time Series Forecasting with LLM-Centric Proxy Denoisingcited by 0
  5. 2026We Need a More Robust Classifier: Dual Causal Learning Empowers Domain-Incremental Time Series Classificationcited by 0
  6. 2026LEFT: Learnable Fusion of Tri-view Tokens for Unsupervised Time Series Anomaly Detectioncited by 0
  7. 2026MLOW: Interpretable Low-Rank Frequency Magnitude Decomposition of Multiple Effects for Time Series Forecastingcited by 0
  8. 2026PAMod: Modeling Cyclical Shifts via Phase-Amplitude Modulation for Non-stationary Time Series Forecastingcited by 0
  9. 2026TopoBrick: Agentic Topology Sampling of Exogenous Variables for Zero-Shot Building IoT Forecastingcited by 0
  10. 2026AirFlow: Context Preserving and Multi-Rate State Modeling for Air Quality Forecastingcited by 0
  11. 2026MetaCaster: Meta-Harness-Optimized Agent for End-to-End Few-Shot Learning of Lightweight Time Series Forecasterscited by 0