mlsanityData to 7 Oct 2026

A Multi-view Multi-task Learning Framework for Multi-variate Time Series Forecasting

arXiv 2109.01657 ↗First posted 2 Sep 2021cs.LGcited by 21

Not on the map. Why?

Abstract (arXiv)

Multi-variate time series (MTS) data is a ubiquitous class of data abstraction in the real world. Any instance of MTS is generated from a hybrid dynamical system and their specific dynamics are usually unknown. The hybrid nature of such a dynamical system is a result of complex external attributes, such as geographic location and time of day, each of which can be categorized into either spatial attributes or temporal attributes. Therefore, there are two fundamental views which can be used to analyze MTS data, namely the spatial view and the temporal view. Moreover, from each of these two views, we can partition the set of data samples of MTS into disjoint forecasting tasks in accordance with their associated attribute values. Then, samples of the same task will manifest similar forthcoming pattern, which is less sophisticated to be predicted in comparison with the original single-view setting. Considering this insight, we propose a novel multi-view multi-task (MVMT) learning framework for MTS forecasting. Instead of being explicitly presented in most scenarios, MVMT information is deeply concealed in the MTS data, which severely hinders the model from capturing it naturally. To this end, we develop two kinds of basic operations, namely task-wise affine transformation and task-wise normalization, respectively. Applying these two operations with prior knowledge on the spatial and temporal view allows the model to adaptively extract MVMT information while predicting. Extensive experiments on three datasets are conducted to illustrate that canonical architectures can be greatly enhanced by the MVMT learning framework in terms of both effectiveness and efficiency. In addition, we design rich case studies to reveal the properties of representations produced at different phases in the entire prediction procedure.

Key facts. Cited by 21 papers in the map (data through 7 Oct 2026). Builds on 36 of its 60 references that are in the map. First posted on arXiv 2 Sep 2021.

Builds on

36 of 60 references.

References · 36

  1. 2017Attention Is All You Needcited by 59,932
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 6,979
  4. 2016Layer Normalizationcited by 5,579
  5. 2020Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecastingcited by 2,141
  6. 2016WaveNet: A Generative Model for Raw Audiocited by 1,710
  7. 2018Group Normalizationcited by 1,514
  8. 2018An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modelingcited by 1,447
  9. 2016Instance Normalization: The Missing Ingredient for Fast Stylizationcited by 1,187
  10. 2017Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecastingcited by 1,125
  11. 2017DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networkscited by 1,004
  12. 2017Modeling Long- and Short-Term Temporal Patterns with Deep Neural Networkscited by 869
24 more
  1. 2019Graph WaveNet for Deep Spatial-Temporal Graph Modelingcited by 820
  2. 2019Enhancing the Locality and Breaking the Memory Bottleneck of Transformer on Time Series Forecastingcited by 606
  3. 2020Connecting the Dots: Multivariate Time Series Forecasting with Graph Neural Networkscited by 563
  4. 2020Adaptive Graph Convolutional Recurrent Network for Traffic Forecastingcited by 525
  5. 2018How Does Batch Normalization Help Optimization?cited by 355
  6. 2016Deep Spatio-Temporal Residual Networks for Citywide Crowd Flows Predictioncited by 330
  7. 2021Spectral Temporal Graph Neural Network for Multivariate Time-series Forecastingcited by 298
  8. 2017A Dual-Stage Attention-Based Recurrent Neural Network for Time Series Predictioncited by 245
  9. 2019Think Globally, Act Locally: A Deep Neural Network Approach to High-Dimensional Time Series Forecastingcited by 161
  10. 2018Understanding Batch Normalizationcited by 157
  11. 2013A Survey on Multi-view Learningcited by 152
  12. 2015Optimal whitening and decorrelationcited by 93
  13. 2019Deep Factors for Forecastingcited by 86
  14. 2019Shape and Time Distortion Loss for Training Deep Time Series Forecasting Modelscited by 66
  15. 2020Do RNN and LSTM have Long Memory?cited by 47
  16. 2019Exploring Interpretable LSTM Neural Networks over Multi-Variable Datacited by 42
  17. 2019Predicting Citywide Crowd Flows in Irregular Regions Using Multi-View Graph Convolutional Networkscited by 40
  18. 2019Joint Modeling of Local and Global Temporal Dynamics for Multivariate Time Series Forecasting with Missing Valuescited by 37
  19. 2017Autoregressive Convolutional Neural Networks for Asynchronous Time Seriescited by 31
  20. 2019Positional Normalizationcited by 28
  21. 2020Pay Attention to Evolution: Time Series Forecasting with Deep Graph-Evolution Learningcited by 22
  22. 2018A Memory-Network Based Solution for Multivariate Time-Series Forecastingcited by 21
  23. 2019Towards Better Forecasting by Fusing Near and Distant Future Visionscited by 14
  24. 2020An Investigation into the Stochasticity of Batch Whiteningcited by 14

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 1; 2023: 7; 2024: 5; 2025: 6; 2026: 2.

21 papers.

Sort:

  1. 2023Exploring Progress in Multivariate Time Series Forecasting: Comprehensive Benchmarking and Heterogeneity Analysiscited by 113
  2. 2023Large Language Model-Aware In-Context Learning for Code Generationcited by 42
  3. 2023Spatial-Temporal-Decoupled Masked Pre-training for Spatiotemporal Forecastingcited by 27
  4. 2023Disentangling Structured Components: Towards Adaptive, Interpretable and Scalable Time Series Forecastingcited by 19
  5. 2025Merlin: Multi-View Representation Learning for Robust Multivariate Time Series Forecasting with Unfixed Missing Ratescited by 19
  6. 2023STAEformer: Spatio-Temporal Adaptive Embedding Makes Vanilla Transformer SOTA for Traffic Forecastingcited by 12
  7. 2024Mitigating Propensity Bias of Large Language Models for Recommender Systemscited by 8
  8. 2023Learning Gaussian Mixture Representations for Tensor Time Series Forecastingcited by 6
  9. 2024Modeling Temporal Dependencies within the Target for Long-Term Time Series Forecastingcited by 6
  10. 2025How Different from the Past? Spatio-Temporal Time Series Forecasting with Self-Supervised Deviation Learningcited by 4
  11. 2022MegaCRN: Meta-Graph Convolutional Recurrent Network for Spatio-Temporal Modelingcited by 3
  12. 2024Prediction of Vessel Arrival Time to Pilotage Area Using Multi-Data Fusion and Deep Learningcited by 3
9 more
  1. 2024Enhancing Spatio-temporal Quantile Forecasting with Curriculum Learning: Lessons Learnedcited by 2
  2. 2024Tackling Data Heterogeneity in Federated Time Series Forecastingcited by 2
  3. 2026Zeus: Towards Tuning-Free Foundation Model for Time Series Analysiscited by 2
  4. 2025Learning Generalized and Flexible Trajectory Models from Omni-Semantic Supervisioncited by 1
  5. 2025Non-collective Calibrating Strategy for Time Series Forecastingcited by 1
  6. 2023Interpretable System Identification and Long-term Prediction on Time-Series Datacited by 0
  7. 2025Accurate and Efficient Multivariate Time Series Forecasting via Offline Clusteringcited by 0
  8. 2025Adaptive Multi-task Learning for Probabilistic Load Forecastingcited by 0
  9. 2026Multi-Task Learning for Sparsely-Labeled Time Series: A Case Study on Cold-Hardiness Modelingcited by 0