Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
119 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Deep Coupling Network For Multivariate Time Series Forecasting (2402.15134v1)

Published 23 Feb 2024 in cs.LG and cs.AI

Abstract: Multivariate time series (MTS) forecasting is crucial in many real-world applications. To achieve accurate MTS forecasting, it is essential to simultaneously consider both intra- and inter-series relationships among time series data. However, previous work has typically modeled intra- and inter-series relationships separately and has disregarded multi-order interactions present within and between time series data, which can seriously degrade forecasting accuracy. In this paper, we reexamine intra- and inter-series relationships from the perspective of mutual information and accordingly construct a comprehensive relationship learning mechanism tailored to simultaneously capture the intricate multi-order intra- and inter-series couplings. Based on the mechanism, we propose a novel deep coupling network for MTS forecasting, named DeepCN, which consists of a coupling mechanism dedicated to explicitly exploring the multi-order intra- and inter-series relationships among time series data concurrently, a coupled variable representation module aimed at encoding diverse variable patterns, and an inference module facilitating predictions through one forward step. Extensive experiments conducted on seven real-world datasets demonstrate that our proposed DeepCN achieves superior performance compared with the state-of-the-art baselines.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (68)
  1. Time Series Using Exponential Smoothing Cells. (2017).
  2. D. Asteriou and S. G. Hall. 2016. ARIMA Models and the Box-Jenkins Methodology. Applied Econometrics.
  3. Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting. In NeurIPS.
  4. An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling. CoRR abs/1803.01271 (2018).
  5. L. E. Baum and T. Petrie. 1966. Statistical Inference for Probabilistic Functions of Finite State Markov Chains. Annals of Mathematical Statistics 37, 6 (1966), 1554–1563.
  6. Representation Learning: A Review and New Perspectives. IEEE Trans. Pattern Anal. Mach. Intell. 35, 8 (2013), 1798–1828.
  7. Autoregressive Convolutional Neural Networks for Asynchronous Time Series. In ICML (Proceedings of Machine Learning Research, Vol. 80). PMLR, 579–588.
  8. Conditional Time Series Forecasting with Convolutional Neural Networks. arXiv (2017).
  9. Spectral Temporal Graph Neural Network for Multivariate Time-series Forecasting. In NeurIPS.
  10. Longbing Cao. 2015. Coupling learning of complex interactions. Inf. Process. Manag. 51, 2 (2015), 167–186.
  11. Deep Modeling Complex Couplings within Financial Markets. In AAAI. AAAI Press, 2518–2524.
  12. N-HiTS: Neural Hierarchical Interpolation for Time Series Forecasting. CoRR abs/2201.12886 (2022).
  13. TAMP-S2GCNets: Coupling Time-Aware Multipersistence Knowledge Representation with Spatio-Supra Graph Convolutional Networks for Time-Series Forecasting. In International Conference on Learning Representations.
  14. Wide & Deep Learning for Recommender Systems. In DLRS@RecSys. ACM, 7–10.
  15. The UCR Time Series Archive. (2018).
  16. Forecasting Big Time Series: Theory and Practice. In KDD. ACM, 3209–3210.
  17. Multi-Horizon Time Series Forecasting with Temporal Attention Learning. In KDD. ACM, 2527–2535.
  18. Temporal Relational Ranking for Stock Prediction. ACM Trans. Inf. Syst. 37, 2 (2019), 27:1–27:30.
  19. DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks. CoRR abs/1704.04110 (2017).
  20. Exploring interpretable LSTM neural networks over multi-variable data. In ICML (Proceedings of Machine Learning Research, Vol. 97). PMLR, 2494–2504.
  21. Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long Short-Term Memory. Neural Comput. 9, 8 (1997), 1735–1780.
  22. Exploring Feature Coupling and Model Coupling for Image Source Identification. IEEE Trans. Inf. Forensics Secur. 13, 12 (2018), 3108–3121.
  23. [Springer Series in Statistics] Forecasting with Exponential Smoothing —— Conventional State Space Models. 10.1007/978-3-540-71918-2, Chapter 13 (2008), 209–227.
  24. CURE: Flexible Categorical Data Representation by Hierarchical Coupling Learning. IEEE Trans. Knowl. Data Eng. 31, 5 (2019), 853–866.
  25. Reformer: The Efficient Transformer. In ICLR.
  26. ImageNet Classification with Deep Convolutional Neural Networks. In NIPS. 1106–1114.
  27. Modeling Long- and Short-Term Temporal Patterns with Deep Neural Networks. In SIGIR. ACM, 95–104.
  28. An Experimental Review on Deep Learning Architectures for Time Series Forecasting. Int. J. Neural Syst. 31, 3 (2021), 2130001:1–2130001:28.
  29. Multi-task Recurrent Neural Networks and Higher-order Markov Random Fields for Stock Price Movement Prediction: Multi-task RNN and Higer-order MRFs for Stock Price Classification. In KDD. ACM, 1141–1151.
  30. J. Li and W. Chen. 2014. Forecasting macroeconomic time series: LASSO-based approaches and their forecast combinations with dynamic factor models. International Journal of Forecasting 30, 4 (2014), 996–1015.
  31. Enhancing the locality and breaking the memory bottleneck of transformer on time series forecasting. Advances in neural information processing systems 32 (2019).
  32. Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting. In ICLR (Poster).
  33. xDeepFM: Combining Explicit and Implicit Feature Interactions for Recommender Systems. In KDD. ACM, 1754–1763.
  34. Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting. CoRR abs/1912.09363 (2019).
  35. Bryan Lim and Stefan Zohren. 2021a. Time-series forecasting with deep learning: a survey. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 379, 2194 (feb 2021), 20200209. https://doi.org/10.1098/rsta.2020.0209
  36. B. Lim and S. Zohren. 2021b. Time-series forecasting with deep learning: a survey. Philosophical Transactions of The Royal Society A Mathematical Physical and Engineering Sciences 379, 2194 (2021), 20200209.
  37. Multi-Task Feature Interaction Learning. In KDD. ACM, 1735–1744.
  38. SCINet: Time Series Modeling and Forecasting with Sample Convolution and Interaction. Thirty-sixth Conference on Neural Information Processing Systems (NeurIPS), 2022 (2022).
  39. Pyraformer: Low-Complexity Pyramidal Attention for Long-Range Time Series Modeling and Forecasting. In ICLR. OpenReview.net.
  40. Time Series Contrastive Learning with Information-Aware Augmentations. In AAAI. AAAI Press, 4534–4542.
  41. Recursive wind speed forecasting based on Hammerstein Auto-Regressive model. Applied Energy 145, may 1 (2015), 191–197.
  42. Nam Nguyen and Brian Quanz. 2021. Temporal Latent Auto-Encoder: A Method for Probabilistic Multivariate Time Series Forecasting. In AAAI. AAAI Press, 9117–9125.
  43. A Time Series is Worth 64 Words: Long-term Forecasting with Transformers. In International Conference on Learning Representations.
  44. N-BEATS: Neural basis expansion analysis for interpretable time series forecasting. arXiv preprint arXiv:1905.10437 (2019).
  45. A nonlinear correlation measure for multivariable data set. Physica D Nonlinear Phenomena 200, 3-4 (2005), 287–295.
  46. A Dual-Stage Attention-Based Recurrent Neural Network for Time Series Prediction. In IJCAI. ijcai.org, 2627–2633.
  47. Deep State Space Models for Time Series Forecasting. In NeurIPS. 7796–7805.
  48. Think Globally, Act Locally: A Deep Neural Network Approach to High-Dimensional Time Series Forecasting. CoRR abs/1905.03806 (2019).
  49. S. Smyl. 2020. A hybrid method of exponential smoothing and recurrent neural networks for time series forecasting. International journal of forecasting 36, 1 (2020), 75–85.
  50. Deep Learning for Time Series Forecasting: A Survey. Big Data 9 (12 2020). https://doi.org/10.1089/big.2020.0159
  51. Attention is All you Need. In NIPS. 5998–6008.
  52. Deep & Cross Network for Ad Click Predictions. In ADKDD@KDD. ACM, 12:1–12:7.
  53. Mark W. Watson. 1993. Vector autoregressions and cointegration. Working Paper Series, Macroeconomic Issues 4 (1993).
  54. William W. S. Wei. 1989. Time series analysis - univariate and multivariate methods. Addison-Wesley.
  55. CoST: Contrastive Learning of Disentangled Seasonal-Trend Representations for Time Series Forecasting. In ICLR. OpenReview.net.
  56. Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting. In NeurIPS. 22419–22430.
  57. Connecting the Dots: Multivariate Time Series Forecasting with Graph Neural Networks. In KDD. 753–763.
  58. Graph WaveNet for Deep Spatial-Temporal Graph Modeling. In IJCAI. 1907–1913.
  59. Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecasting. In IJCAI. ijcai.org, 3634–3640.
  60. Temporal Regularized Matrix Factorization for High-dimensional Time Series Prediction. In NIPS. 847–855.
  61. TS2Vec: Towards Universal Representation of Time Series. In AAAI. AAAI Press, 8980–8987.
  62. Are Transformers Effective for Time Series Forecasting? Proceedings of the AAAI Conference on Artificial Intelligence.
  63. Stock Price Prediction via Discovering Multi-Frequency Trading Patterns. In KDD. 2141–2149.
  64. Neural Time-Aware Sequential Recommendation by Jointly Modeling Preference Dynamics and Explicit Feature Couplings. IEEE Trans. Neural Networks Learn. Syst. 33, 10 (2022), 5125–5137.
  65. Forecasting Fine-Grained Air Quality Based on Big Data. In KDD. ACM, 2267–2276.
  66. CAN: Revisiting Feature Co-Action for Click-Through Rate Prediction. CoRR abs/2011.05625 (2020).
  67. Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting. In AAAI. AAAI Press, 11106–11115.
  68. Unsupervised Heterogeneous Coupling Learning for Categorical Representation. IEEE Trans. Pattern Anal. Mach. Intell. 44, 1 (2022), 533–549.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (7)
  1. Kun Yi (25 papers)
  2. Qi Zhang (785 papers)
  3. Hui He (38 papers)
  4. Kaize Shi (9 papers)
  5. Liang Hu (64 papers)
  6. Ning An (29 papers)
  7. Zhendong Niu (10 papers)
Citations (5)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now