pith. sign in

arxiv: 2605.22540 · v1 · pith:KJXX6MUTnew · submitted 2026-05-21 · 💻 cs.CE · cs.AI

Dynamic Hypergraph Representation Learning for Multivariate Time Series without Prior Knowledge

Marco Gregnanin , Johannes De Smedt , Giorgio Gnecco , Maurizio Parton This is my paper

Pith reviewed 2026-05-22 01:43 UTC · model grok-4.3

classification 💻 cs.CE cs.AI
keywords hypergraph representation learningmultivariate time seriescommunity detectiondynamic hypergraphsattention mechanismtime series predictionhigher-order relationshipsclique expansion
0
0 comments X

The pith

A method builds dynamic hypergraphs from multivariate time series data without any prior knowledge of their structure.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a technique to create hypergraph structures from time series data that captures relationships involving multiple variables at once. It starts by applying community detection to the raw data, uses an attention mechanism to pick relevant groups, and then forms hyperedges by connecting all members of each group. These hypergraphs power a specialized network called DHACN to forecast the time series values. A sympathetic reader would care because this removes the need for expert-defined connections, potentially making modeling of complex systems more accessible and accurate.

Core claim

The authors introduce a model that applies community detection to multivariate time series, selects communities with an attention mechanism, and expands them into hyperedges using clique expansion to form a dynamic hypergraph. This hypergraph is then processed by DHACN to predict future values in the time series, all without relying on predefined graph structures.

What carries the argument

Community detection on raw time series followed by attention-based selection and clique expansion to form hyperedges, processed by the Dynamic Hypergraph Attention Convolution Network (DHACN).

If this is right

  • The approach enables hypergraph-based modeling of multivariate time series in settings where structural information is unavailable.
  • Higher-order relationships among variables can be used directly for forecasting without first reducing them to pairwise links.
  • The dynamic construction allows the hypergraph to update as new time series observations arrive.
  • Predictions benefit from the richer connectivity provided by hyperedges compared with standard graph methods.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same community-to-hyperedge pipeline could be tested on domains such as financial returns or sensor networks to see whether it surfaces interpretable multi-variable groups.
  • One could measure how much the attention step improves hyperedge quality by ablating it and comparing forecast error.
  • If the method generalizes, it might reduce reliance on hand-crafted features in any time series task where group-wise dependencies matter.

Load-bearing premise

Community detection applied directly to the raw time series, followed by attention-based selection and clique expansion, produces hyperedges that faithfully represent the underlying higher-order dynamics of the system.

What would settle it

Test the constructed hyperedges and prediction accuracy on a synthetic multivariate time series dataset engineered with known higher-order interactions and check whether the detected communities recover those interactions.

Figures

Figures reproduced from arXiv: 2605.22540 by Giorgio Gnecco, Johannes De Smedt, Marco Gregnanin, Maurizio Parton.

Figure 1
Figure 1. Figure 1: An overall pipeline of the proposed approach, showing all the steps needed to arrive at the [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Proposed Hypergraph Neural Networks Architecture. [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
read the original abstract

Hypergraphs have the capacity to capture higher-dimensional relationships among entities across various domains, making them a subject of growing interest within the research community for understanding the structure and dynamics of complex systems. However, a key challenge is the derivation of hypergraph representations from time series data in situations where the structure of the hypergraph is limited or absent. In this study, we propose a model that constructs a dynamic hypergraph representation for multivariate time series without relying on prior knowledge of the data. This is achieved by applying community detection to the time series and transforming the resulting communities, obtained through an attention mechanism, into a hypergraph using a clique-based technique. Hypergraph representations are derived from different time series datasets, and the resulting hypergraphs are then used by a Dynamic Hypergraph Attention Convolution Network (DHACN) for multivariate time series predictions. This research advances the field of hypergraph representation by introducing a novel approach that is better suited to uncover high-order relationships without prior knowledge.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript proposes a model to construct dynamic hypergraph representations from multivariate time series data without prior knowledge. Community detection is applied to the time series, communities are selected via an attention mechanism, and the results are transformed into hyperedges using a clique-based expansion. The resulting hypergraphs are fed into a Dynamic Hypergraph Attention Convolution Network (DHACN) for multivariate time series prediction.

Significance. If the hyperedges constructed via community detection and clique expansion demonstrably capture higher-order interactions beyond pairwise correlations, the approach could offer a practical route to hypergraph-based modeling of time series in domains where structural priors are unavailable. The combination of community detection, attention, and dynamic convolution is a reasonable direction, but its value depends on showing that the derived hypergraphs improve predictions in a manner not achievable by standard graph methods.

major comments (2)
  1. [Methods] Methods section: The central construction applies community detection directly to the multivariate time series (presumably after forming a pairwise similarity graph), selects communities via attention, and expands them into hyperedges via cliques. For the claim of uncovering high-order relationships without prior knowledge to hold, these communities must correspond to multi-way interactions rather than merely reflecting pairwise correlations already present in the data. The paper provides no explicit validation (e.g., against synthetic hypergraph benchmarks with known higher-order structure) that the resulting hyperedges add information beyond what a standard graph would capture.
  2. [Experiments] Experiments section: No equations, experimental results, ablation studies, or error bars are visible, so it is impossible to verify whether the described steps actually support the performance claims. In particular, comparisons to graph-based baselines and tests on data with known higher-order ground truth are required to substantiate the advance over existing hypergraph representation techniques.
minor comments (1)
  1. [Abstract] Abstract: The abstract states that hypergraphs are derived from different time series datasets but does not name the datasets or report quantitative gains; adding these details would improve clarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive comments on our manuscript. We address each major comment below and indicate the changes planned for the revised version.

read point-by-point responses

  1. Referee: [Methods] Methods section: The central construction applies community detection directly to the multivariate time series (presumably after forming a pairwise similarity graph), selects communities via attention, and expands them into hyperedges via cliques. For the claim of uncovering high-order relationships without prior knowledge to hold, these communities must correspond to multi-way interactions rather than merely reflecting pairwise correlations already present in the data. The paper provides no explicit validation (e.g., against synthetic hypergraph benchmarks with known higher-order structure) that the resulting hyperedges add information beyond what a standard graph would capture.

    Authors: We agree that explicit validation is required to substantiate that the hyperedges capture higher-order interactions. The current construction begins with a pairwise similarity graph before community detection and clique expansion; while the expansion step is designed to encode multi-way relations within detected communities, we will add a dedicated paragraph in the Methods section explaining this motivation with supporting references. We will also incorporate new experiments on synthetic hypergraphs with known higher-order structure and direct comparisons against pairwise graph baselines using the same community detection step but without clique expansion. revision: yes

  2. Referee: [Experiments] Experiments section: No equations, experimental results, ablation studies, or error bars are visible, so it is impossible to verify whether the described steps actually support the performance claims. In particular, comparisons to graph-based baselines and tests on data with known higher-order ground truth are required to substantiate the advance over existing hypergraph representation techniques.

    Authors: We acknowledge that the experimental details, equations, quantitative results, ablations, and error bars were insufficiently prominent or referenced in the submitted version. In the revision we will expand the Experiments section to include the full set of DHACN equations, all performance tables with error bars from multiple runs, ablation studies on the attention and community modules, comparisons to graph-based baselines, and additional tests on synthetic data possessing known higher-order ground truth. revision: yes

Circularity Check

0 steps flagged

No significant circularity in the proposed hypergraph construction pipeline.

full rationale

The paper presents a forward methodological pipeline: community detection is applied to the multivariate time series, communities are selected via an attention mechanism, and the result is expanded into hyperedges via a clique-based technique to form a dynamic hypergraph. This hypergraph is then fed into the DHACN model for time-series prediction. The derivation is described as a constructive procedure that does not reduce any claimed prediction or first-principles result to its own inputs by definition, nor does it rely on load-bearing self-citations or uniqueness theorems imported from prior author work. No fitted parameters are renamed as independent predictions, and the central claim of uncovering higher-order relationships rests on the explicit steps of the pipeline rather than circular re-use of the target quantity. The construction is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The method rests on the domain assumption that standard community detection applied to time series yields groups that correspond to meaningful higher-order relations, plus the modeling choice that clique expansion is an appropriate way to turn communities into hyperedges. No free parameters or invented entities are named in the abstract.

axioms (1)
  • domain assumption Community detection on multivariate time series produces groups that reflect higher-order relationships without prior structural knowledge
    Invoked when the paper states that hypergraph representations are derived by applying community detection to the time series

pith-pipeline@v0.9.0 · 5704 in / 1354 out tokens · 32701 ms · 2026-05-22T01:43:36.749812+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

52 extracted references · 52 canonical work pages · 3 internal anchors

  1. [1]

    Optuna: A next-generation hyperparameter optimization framework

    Takuya Akiba, Shotaro Sano, Toshihiko Yanase, Takeru Ohta, and Masanori Koyama. Optuna: A next-generation hyperparameter optimization framework. InProceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, pages 2623–2631, 2019

    work page 2019

  2. [2]

    A Survey on Hypergraph Representation Learning.ACM Computing Surveys, 56(1):1–38, 2023

    Alessia Antelmi, Gennaro Cordasco, Mirko Polato, Vittorio Scarano, Carmine Spagnuolo, and Dingqi Yang. A Survey on Hypergraph Representation Learning.ACM Computing Surveys, 56(1):1–38, 2023

    work page 2023

  3. [3]

    Hypergraph Convolution and Hypergraph Attention

    Song Bai, Feihu Zhang, and Philip HS Torr. Hypergraph Convolution and Hypergraph Attention. Pattern Recognition, 110:107637, 2021

    work page 2021

  4. [4]

    Cambridge University Press, Cambridge, 2016

    Albert-László Barabási and Márton Pósfai.Network Science. Cambridge University Press, Cambridge, 2016

    work page 2016

  5. [5]

    An Introduction

    Alain Bretto.Hypergraph Theory. An Introduction. Springer International Publishing, 2013

    work page 2013

  6. [6]

    Appliances Energy Prediction

    Luis Candanedo. Appliances Energy Prediction. UCI Machine Learning Repository, 2017. DOI: https://doi.org/10.24432/C5VC8G

    work page doi:10.24432/c5vc8g 2017

  7. [7]

    Spectral Temporal Graph Neural Network for Multivariate Time-Series Forecasting.Advances in neural information processing systems, 33:17766–17778, 2020

    Defu Cao, Yujing Wang, Juanyong Duan, Ce Zhang, Xia Zhu, Congrui Huang, Yunhai Tong, Bixiong Xu, Jing Bai, Jie Tong, et al. Spectral Temporal Graph Neural Network for Multivariate Time-Series Forecasting.Advances in neural information processing systems, 33:17766–17778, 2020

    work page 2020

  8. [8]

    On the Properties of Neural Machine Translation: Encoder-decoder Approaches, 2014

    Kyunghyun Cho, Bart van Merrienboer, Dzmitry Bahdanau, and Yoshua Bengio. On the Properties of Neural Machine Translation: Encoder-decoder Approaches, 2014

    work page 2014

  9. [9]

    Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)

    Djork-Arné Clevert, Thomas Unterthiner, and Sepp Hochreiter. Fast and accurate deep network learning by exponential linear units (elus), 2016.https://arxiv.org/abs/1511.07289

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  10. [10]

    Random Matrix Theory.Acta Numerica, 14:233–297, 2005

    Alan Edelman and N Raj Rao. Random Matrix Theory.Acta Numerica, 14:233–297, 2005

    work page 2005

  11. [11]

    Hypergraph Neural Networks

    Yifan Feng, Haoxuan You, Zizhao Zhang, Rongrong Ji, and Yue Gao. Hypergraph Neural Networks. InProceedings of the AAAI conference on artificial intelligence, volume 33, pages 3558–3565, 2019

    work page 2019

  12. [12]

    Time Series Clustering via Community Detection in Networks.Information Sciences, 326:227–242, 2016

    Leonardo N Ferreira and Liang Zhao. Time Series Clustering via Community Detection in Networks.Information Sciences, 326:227–242, 2016

    work page 2016

  13. [13]

    Community Detection in Graphs.Physics Reports, 486(3-5):75–174, 2010

    Santo Fortunato. Community Detection in Graphs.Physics Reports, 486(3-5):75–174, 2010

    work page 2010

  14. [14]

    HGNN+: General Hypergraph Neural Networks.IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(3):3181–3199, 2022

    Yue Gao, Yifan Feng, Shuyi Ji, and Rongrong Ji. HGNN+: General Hypergraph Neural Networks.IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(3):3181–3199, 2022

    work page 2022

  15. [15]

    3-d Object Retrieval and Recognition with Hypergraph Analysis.IEEE transactions on image processing, 21(9):4290– 4303, 2012

    Yue Gao, Meng Wang, Dacheng Tao, Rongrong Ji, and Qionghai Dai. 3-d Object Retrieval and Recognition with Hypergraph Analysis.IEEE transactions on image processing, 21(9):4290– 4303, 2012

    work page 2012

  16. [16]

    Gregnanin, J

    M. Gregnanin, J. De Smedt, G. Gnecco, and M. Parton. Signature-Based Community Detection for Time Series. InComplex Networks & Their Applications XII, volume 1142, pages 146–158, Cham, 2024. Springer Nature Switzerland. 10

    work page 2024

  17. [17]

    The statistical significance of the inclusion of graph neural networks in the financial time series forecasting problem, 2026

    Marco Gregnanin, Johannes De Smedt, Giorgio Gnecco, and Maurizio Parton. The statistical significance of the inclusion of graph neural networks in the financial time series forecasting problem, 2026

    work page 2026

  18. [18]

    Long Short-Term Memory.Neural computation, 9(8):1735–1780, 1997

    Sepp Hochreiter and Jürgen Schmidhuber. Long Short-Term Memory.Neural computation, 9(8):1735–1780, 1997

    work page 1997

  19. [19]

    Unignn: A Unified Framework for Graph and Hypergraph Neural Networks.arXiv preprint arXiv:2105.00956, 2021

    Jing Huang and Jie Yang. Unignn: A Unified Framework for Graph and Hypergraph Neural Networks.arXiv preprint arXiv:2105.00956, 2021

    work page arXiv 2021

  20. [20]

    Learning Hypergraph- Regularized Attribute Predictors

    Sheng Huang, Mohamed Elhoseiny, Ahmed Elgammal, and Dan Yang. Learning Hypergraph- Regularized Attribute Predictors. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 409–417, 2015

    work page 2015

  21. [21]

    Dynamic Hypergraph Neural Networks

    Jianwen Jiang, Yuxuan Wei, Yifan Feng, Jingxuan Cao, and Yue Gao. Dynamic Hypergraph Neural Networks. In28th International Joint Conference on Artificial Intelligence (IJCAI), pages 2635–2641, 2019

    work page 2019

  22. [22]

    Robust l2- Hypergraph and its Applications.Information Sciences, 501:708–723, 2019

    Taisong Jin, Zhengtao Yu, Yue Gao, Shengxiang Gao, Xiaoshuai Sun, and Cuihua Li. Robust l2- Hypergraph and its Applications.Information Sciences, 501:708–723, 2019

    work page 2019

  23. [23]

    Dynamic Hypergraph Neural Networks Based On Key Hyperedges

    Xiaojun Kang, Xinchuan Li, Hong Yao, Dan Li, Bo Jiang, Xiaoyue Peng, Tiejun Wu, Shihua Qi, and Lijun Dong. Dynamic Hypergraph Neural Networks Based On Key Hyperedges. Information Sciences, 616:37–51, 2022

    work page 2022

  24. [24]

    Adam: A Method for Stochastic Optimization

    Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization.arXiv preprint arXiv:1412.6980, 2014

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  25. [25]

    Noise Dressing of Financial Correlation Matrices.Physical review letters, 83(7):1467, 1999

    Laurent Laloux, Pierre Cizeau, Jean-Philippe Bouchaud, and Marc Potters. Noise Dressing of Financial Correlation Matrices.Physical review letters, 83(7):1467, 1999

    work page 1999

  26. [26]

    Random Natrix Theory and Financial Correlations.International Journal of Theoretical and Applied Finance, 3(03):391–397, 2000

    Laurent Laloux, Pierre Cizeau, Marc Potters, and Jean-Philippe Bouchaud. Random Natrix Theory and Financial Correlations.International Journal of Theoretical and Applied Finance, 3(03):391–397, 2000

    work page 2000

  27. [27]

    News Recommendation Via Hypergraph Learning: Encapsulation of User Behavior and News Content

    Lei Li and Tao Li. News Recommendation Via Hypergraph Learning: Encapsulation of User Behavior and News Content. InProceedings of the sixth ACM international conference on Web search and data mining, pages 305–314, 2013

    work page 2013

  28. [28]

    Directed Hypergraph Attention Network for Traffic Forecasting.IET Intelligent Transport Systems, 16(1):85–98, 2022

    Xiaoyi Luo, Jiaheng Peng, and Jun Liang. Directed Hypergraph Attention Network for Traffic Forecasting.IET Intelligent Transport Systems, 16(1):85–98, 2022

    work page 2022

  29. [29]

    Community Detection for Correlation Matrices

    Mel MacMahon and Diego Garlaschelli. Community Detection for Correlation Matrices. Physical Review X, 5(2), apr 2015

    work page 2015

  30. [30]

    Cambridge University Press, 1999

    Rosario N Mantegna and H Eugene Stanley.Introduction to Econophysics: Correlations and Complexity in Finance. Cambridge University Press, 1999

    work page 1999

  31. [31]

    Elsevier, 2004

    Madan Lal Mehta.Random Matrices, volume 142. Elsevier, 2004

    work page 2004

  32. [32]

    Finding and Evaluating Community Structure in Networks.Physical review E, 69(2):026113, 2004

    Mark EJ Newman and Michelle Girvan. Finding and Evaluating Community Structure in Networks.Physical review E, 69(2):026113, 2004

    work page 2004

  33. [33]

    K-Nearest Neighbor.Scholarpedia, 4(2):1883, 2009

    Leif E Peterson. K-Nearest Neighbor.Scholarpedia, 4(2):1883, 2009

    work page 2009

  34. [34]

    Random matrix approach to cross correlations in financial data

    Vasiliki Plerou, Parameswaran Gopikrishnan, Bernd Rosenow, Luís A Nunes Amaral, Thomas Guhr, and H Eugene Stanley. Random matrix approach to cross correlations in financial data. Physical Review E, 65(6):066126, 2002

    work page 2002

  35. [35]

    Universal and Nonuniversal Properties of Cross Correlations in Financial Time Series.Physical review letters, 83(7):1471, 1999

    Vasiliki Plerou, Parameswaran Gopikrishnan, Bernd Rosenow, Luís A Nunes Amaral, and H Eugene Stanley. Universal and Nonuniversal Properties of Cross Correlations in Financial Time Series.Physical review letters, 83(7):1471, 1999

    work page 1999

  36. [36]

    Financial Applications of Random Matrix Theory: Old Laces and New Pieces

    Marc Potters, Jean-Philippe Bouchaud, and Laurent Laloux. Financial applications of random matrix theory: Old laces and new pieces.arXiv preprint physics/0507111, 2005. 11

    work page internal anchor Pith review Pith/arXiv arXiv 2005

  37. [37]

    Stock Selection via Spatiotemporal Hypergraph Attention Network: A Learning to Rank Approach

    Ramit Sawhney, Shivam Agarwal, Arnav Wadhwa, Tyler Derr, and Rajiv Ratn Shah. Stock Selection via Spatiotemporal Hypergraph Attention Network: A Learning to Rank Approach. Proceedings of the AAAI Conference on Artificial Intelligence, 35(1):497–504, May 2021

    work page 2021

  38. [38]

    The Graph Neural Network model.IEEE transactions on neural networks, 20(1):61–80, 2008

    Franco Scarselli, Marco Gori, Ah Chung Tsoi, Markus Hagenbuchner, and Gabriele Monfardini. The Graph Neural Network model.IEEE transactions on neural networks, 20(1):61–80, 2008

    work page 2008

  39. [39]

    John Wiley & Sons, 2010

    Sitabhra Sinha, Arnab Chatterjee, Anirban Chakraborti, and Bikas K Chakrabarti.Econophysics: an Introduction. John Wiley & Sons, 2010

    work page 2010

  40. [40]

    A tool for filtering information in complex systems.Proceedings of the National Academy of Sciences, 102(30):10421–10426, 2005

    Michele Tumminello, Tomaso Aste, Tiziana Di Matteo, and Rosario N Mantegna. A tool for filtering information in complex systems.Proceedings of the National Academy of Sciences, 102(30):10421–10426, 2005

    work page 2005

  41. [41]

    Random matrix theory analysis of cross correlations in financial markets.Physical Review E—Statistical, Nonlinear, and Soft Matter Physics, 70(2):026110, 2004

    Akihiko Utsugi, Kazusumi Ino, and Masaki Oshikawa. Random matrix theory analysis of cross correlations in financial markets.Physical Review E—Statistical, Nonlinear, and Soft Matter Physics, 70(2):026110, 2004

    work page 2004

  42. [42]

    Attention is All You Need.Advances in neural information processing systems, 30, 2017

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is All You Need.Advances in neural information processing systems, 30, 2017

    work page 2017

  43. [43]

    Air Quality

    Saverio Vito. Air Quality. UCI Machine Learning Repository, 2008. DOI: https://doi.org/10.24432/C59K5F

    work page doi:10.24432/c59k5f 2008

  44. [44]

    Multitask Hypergraph Convolutional Networks: A Heterogeneous Traffic Prediction Framework.IEEE Transactions on Intelligent Transportation Systems, 23(10):18557–18567, 2022

    Jingcheng Wang, Yong Zhang, Lixun Wang, Yongli Hu, Xinglin Piao, and Baocai Yin. Multitask Hypergraph Convolutional Networks: A Heterogeneous Traffic Prediction Framework.IEEE Transactions on Intelligent Transportation Systems, 23(10):18557–18567, 2022

    work page 2022

  45. [45]

    Dynamic Hypergraph Structure Learning for Multivariate Time Series Forecasting.IEEE Transactions on Big Data, 2024

    Shun Wang, Yong Zhang, Xuanqi Lin, Yongli Hu, Qingming Huang, and Baocai Yin. Dynamic Hypergraph Structure Learning for Multivariate Time Series Forecasting.IEEE Transactions on Big Data, 2024

    work page 2024

  46. [46]

    Hongjie Xia, Huijie Ao, Long Li, Yu Liu, Sen Liu, Guangnan Ye, and Hongfeng Chai. Ci-sthpan: Pre-Trained Attention Network for Stock Selection with Channel-Independent Spatio-Temporal Hypergraph.Proceedings of the AAAI Conference on Artificial Intelligence, 38(8):9187–9195, Mar. 2024

    work page 2024

  47. [47]

    Distinguishing Chaotic Time Series From Noise: A Random Matrix Approach.Communications in Nonlinear Science and Numerical Simulation, 44:284–291, 2017

    Bin Ye, Jianxing Chen, Chen Ju, Huijun Li, and Xuesong Wang. Distinguishing Chaotic Time Series From Noise: A Random Matrix Approach.Communications in Nonlinear Science and Numerical Simulation, 44:284–291, 2017

    work page 2017

  48. [48]

    Hypergraph Convolutional Recurrent Neural Network

    Jaehyuk Yi and Jinkyoo Park. Hypergraph Convolutional Recurrent Neural Network. In Proceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining, pages 3366–3376, 2020

    work page 2020

  49. [49]

    Dynamic Hypergraph Convolutional Network

    Nan Yin, Fuli Feng, Zhigang Luo, Xiang Zhang, Wenjie Wang, Xiao Luo, Chong Chen, and Xian-Sheng Hua. Dynamic Hypergraph Convolutional Network. In2022 IEEE 38th International Conference on Data Engineering (ICDE), pages 1621–1634. IEEE, 2022

    work page 2022

  50. [50]

    Nan Yin, Li Shen, Huan Xiong, Bin Gu, Chong Chen, Xian-Sheng Hua, Siwei Liu, and Xiao Luo. Messages are Never Propagated Alone: Collaborative Hypergraph Neural Network for Time-Series Forecasting.IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(4):2333–2347, 2024

    work page 2024

  51. [51]

    Lipton, Mu Li, and Alexander J

    Aston Zhang, Zachary C. Lipton, Mu Li, and Alexander J. Smola.Dive into Deep Learning. Cambridge University Press, 2023.https://D2L.ai

    work page 2023

  52. [52]

    Dynamic Hypergraph Structure Learning

    Zizhao Zhang, Haojie Lin, Yue Gao, and KLISS BNRist. Dynamic Hypergraph Structure Learning. In27th International Joint Conference on Artificial Intelligence (IJCAI), pages 3162–3169, 2018. 12 A Random Matrix Theory We consider the Random Matrix Theory (RMT) as principal approach for filtering similarity matrices. By filtering, we refer to the process of r...

    work page 2018