mlsanityData to 7 Oct 2026

Few-Shot Forecasting of Time-Series with Heterogeneous Channels

arXiv 2204.03456 ↗First posted 7 Apr 2022cs.LGcited by 7

Not on the map. Why?

Abstract (arXiv)

Learning complex time series forecasting models usually requires a large amount of data, as each model is trained from scratch for each task/data set. Leveraging learning experience with similar datasets is a well-established technique for classification problems called few-shot classification. However, existing approaches cannot be applied to time-series forecasting because i) multivariate time-series datasets have different channels and ii) forecasting is principally different from classification. In this paper we formalize the problem of few-shot forecasting of time-series with heterogeneous channels for the first time. Extending recent work on heterogeneous attributes in vector data, we develop a model composed of permutation-invariant deep set-blocks which incorporate a temporal embedding. We assemble the first meta-dataset of 40 multivariate time-series datasets and show through experiments that our model provides a good generalization, outperforming baselines carried over from simpler scenarios that either fail to learn across tasks or miss temporal information.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 38 of its 60 references that are in the map. First posted on arXiv 7 Apr 2022.

Builds on

38 of 60 references.

References · 38

  1. 2017Attention Is All You Needcited by 59,932
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  4. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
  5. 2017Prototypical Networks for Few-shot Learningcited by 3,001
  6. 2016Xception: Deep Learning with Depthwise Separable Convolutionscited by 2,320
  7. 2016Matching Networks for One Shot Learningcited by 2,246
  8. 2020Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecastingcited by 2,141
  9. 2018DARTS: Differentiable Architecture Searchcited by 1,782
  10. 2020Reformer: The Efficient Transformercited by 1,658
  11. 2020Linformer: Self-Attention with Linear Complexitycited by 1,462
  12. 2021MLP-Mixer: An all-MLP Architecture for Visioncited by 1,441
26 more
  1. 2020Sharpness-Aware Minimization for Efficiently Improving Generalizationcited by 1,404
  2. 2017Deep Setscited by 1,394
  3. 2017DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networkscited by 1,004
  4. 2018On First-Order Meta-Learning Algorithmscited by 1,003
  5. 2019N-BEATS: Neural basis expansion analysis for interpretable time series forecastingcited by 744
  6. 2021CoAtNet: Marrying Convolution and Attention for All Data Sizescited by 566
  7. 2018Meta-Learning with Latent Embedding Optimizationcited by 476
  8. 2020Time Series Forecasting With Deep Learning: A Surveycited by 457
  9. 2018The UCR Time Series Archivecited by 421
  10. 2018TADAM: Task dependent adaptive metric for improved few-shot learningcited by 408
  11. 2016Time Series Classification from Scratch with Deep Neural Networks: A Strong Baselinecited by 358
  12. 2018Invariant and Equivariant Graph Networkscited by 339
  13. 2018The UEA multivariate time series classification archive, 2018cited by 330
  14. 2017Meta Networkscited by 268
  15. 2021Monash Time Series Forecasting Archivecited by 256
  16. 2017A Dual-Stage Attention-Based Recurrent Neural Network for Time Series Predictioncited by 245
  17. 2019Cross Attention Network for Few-shot Classificationcited by 218
  18. 2018Meta-Reinforcement Learning of Structured Exploration Strategiescited by 128
  19. 2020Meta-learning framework with applications to zero-shot time-series forecastingcited by 70
  20. 2019Meta-Learning for Few-Shot Time Series Classificationcited by 23
  21. 2020Few-shot Learning for Time-series Forecastingcited by 15
  22. 2020Interpretable Time-series Classification on Few-shot Samplescited by 11
  23. 2021Yformer: U-Net Inspired Transformer Architecture for Far Horizon Time Series Forecastingcited by 11
  24. 2021Multimodal Meta-Learning for Time Series Regressioncited by 6
  25. 2019Chameleon: Learning Model Initializations Across Tasks With Different Schemascited by 4
  26. 2019HIDRA: Head Initialization across Dynamic targets for Robust Architecturescited by 3

Built on by

Cited by year

Citing papers per year: 2022: 1; 2023: 2; 2024: 3; 2025: 1.

7 papers.

  1. 2023A Survey on Time-Series Pre-Trained Modelscited by 52
  2. 2024A Survey of Few-Shot Learning for Biomedical Time Seriescited by 7
  3. 2022Few-shot human motion prediction for heterogeneous sensorscited by 2
  4. 2025MetaEformer: Unveiling and Leveraging Meta-patterns for Complex and Dynamic Systems Load Forecastingcited by 2
  5. 2024Off-the-shelf ChatGPT is a Good Few-shot Human Motion Predictorcited by 1
  6. 2024Few-Shot Load Forecasting Under Data Scarcity in Smart Grids: A Meta-Learning Approachcited by 1
  7. 2023Forecasting Early with Meta Learningcited by 0