IDEAS home Printed from https://ideas.repec.org/p/arx/papers/2201.08326.html
   My bibliography  Save this paper

Learning with latent group sparsity via heat flow dynamics on networks

Author

Listed:
  • Subhroshekhar Ghosh
  • Soumendu Sundar Mukherjee

Abstract

Group or cluster structure on explanatory variables in machine learning problems is a very general phenomenon, which has attracted broad interest from practitioners and theoreticians alike. In this work we contribute an approach to learning under such group structure, that does not require prior information on the group identities. Our paradigm is motivated by the Laplacian geometry of an underlying network with a related community structure, and proceeds by directly incorporating this into a penalty that is effectively computed via a heat flow-based local network dynamics. In fact, we demonstrate a procedure to construct such a network based on the available data. Notably, we dispense with computationally intensive pre-processing involving clustering of variables, spectral or otherwise. Our technique is underpinned by rigorous theorems that guarantee its effective performance and provide bounds on its sample complexity. In particular, in a wide range of settings, it provably suffices to run the heat flow dynamics for time that is only logarithmic in the problem dimensions. We explore in detail the interfaces of our approach with key statistical physics models in network science, such as the Gaussian Free Field and the Stochastic Block Model. We validate our approach by successful applications to real-world data from a wide array of application domains, including computer science, genetics, climatology and economics. Our work raises the possibility of applying similar diffusion-based techniques to classical learning tasks, exploiting the interplay between geometric, dynamical and stochastic structures underlying the data.

Suggested Citation

  • Subhroshekhar Ghosh & Soumendu Sundar Mukherjee, 2022. "Learning with latent group sparsity via heat flow dynamics on networks," Papers 2201.08326, arXiv.org.
  • Handle: RePEc:arx:papers:2201.08326
    as

    Download full text from publisher

    File URL: http://arxiv.org/pdf/2201.08326
    File Function: Latest version
    Download Restriction: no
    ---><---

    References listed on IDEAS

    as
    1. Frédéric Lavancier & Jesper Møller & Ege Rubak, 2015. "Determinantal point process models and statistical inference," Journal of the Royal Statistical Society Series B, Royal Statistical Society, vol. 77(4), pages 853-877, September.
    2. Vivien Marx, 2013. "The big challenges of big data," Nature, Nature, vol. 498(7453), pages 255-260, June.
    3. Ming Yuan & Yi Lin, 2006. "Model selection and estimation in regression with grouped variables," Journal of the Royal Statistical Society Series B, Royal Statistical Society, vol. 68(1), pages 49-67, February.
    Full references (including those not matched with items on IDEAS)

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Tutz, Gerhard & Pößnecker, Wolfgang & Uhlmann, Lorenz, 2015. "Variable selection in general multinomial logit models," Computational Statistics & Data Analysis, Elsevier, vol. 82(C), pages 207-222.
    2. Lin Zhu & Xiantao Liu & Sha He & Jun Shi & Ming Pang, 2015. "Keywords co-occurrence mapping knowledge domain research base on the theory of Big Data in oil and gas industry," Scientometrics, Springer;Akadémiai Kiadó, vol. 105(1), pages 249-260, October.
    3. Zhaoxing Gao & Ruey S. Tsay, 2021. "Divide-and-Conquer: A Distributed Hierarchical Factor Approach to Modeling Large-Scale Time Series Data," Papers 2103.14626, arXiv.org.
    4. Jun Yan & Jian Huang, 2012. "Model Selection for Cox Models with Time-Varying Coefficients," Biometrics, The International Biometric Society, vol. 68(2), pages 419-428, June.
    5. Zhang, Yi & Huang, Ying & Porter, Alan L. & Zhang, Guangquan & Lu, Jie, 2019. "Discovering and forecasting interactions in big data research: A learning-enhanced bibliometric study," Technological Forecasting and Social Change, Elsevier, vol. 146(C), pages 795-807.
    6. Ye, Ya-Fen & Shao, Yuan-Hai & Deng, Nai-Yang & Li, Chun-Na & Hua, Xiang-Yu, 2017. "Robust Lp-norm least squares support vector regression with feature selection," Applied Mathematics and Computation, Elsevier, vol. 305(C), pages 32-52.
    7. Guillaume Sagnol & Edouard Pauwels, 2019. "An unexpected connection between Bayes A-optimal designs and the group lasso," Statistical Papers, Springer, vol. 60(2), pages 565-584, April.
    8. Diego Vidaurre & Concha Bielza & Pedro Larrañaga, 2013. "A Survey of L1 Regression," International Statistical Review, International Statistical Institute, vol. 81(3), pages 361-387, December.
    9. Bakalli, Gaetan & Guerrier, Stéphane & Scaillet, Olivier, 2023. "A penalized two-pass regression to predict stock returns with time-varying risk premia," Journal of Econometrics, Elsevier, vol. 237(2).
    10. Shuichi Kawano, 2014. "Selection of tuning parameters in bridge regression models via Bayesian information criterion," Statistical Papers, Springer, vol. 55(4), pages 1207-1223, November.
    11. Peng, Heng & Lu, Ying, 2012. "Model selection in linear mixed effect models," Journal of Multivariate Analysis, Elsevier, vol. 109(C), pages 109-129.
    12. Yize Zhao & Matthias Chung & Brent A. Johnson & Carlos S. Moreno & Qi Long, 2016. "Hierarchical Feature Selection Incorporating Known and Novel Biological Information: Identifying Genomic Features Related to Prostate Cancer Recurrence," Journal of the American Statistical Association, Taylor & Francis Journals, vol. 111(516), pages 1427-1439, October.
    13. G. Aneiros & P. Vieu, 2016. "Sparse nonparametric model for regression with functional covariate," Journal of Nonparametric Statistics, Taylor & Francis Journals, vol. 28(4), pages 839-859, October.
    14. Tan Guo & Lei Zhang & Xiaoheng Tan & Liu Yang & Zhiwei Guo & Fupeng Wei, 2019. "CoLR: Classification-Oriented Local Representation for Image Recognition," Complexity, Hindawi, vol. 2019, pages 1-17, June.
    15. He Jiang, 2023. "Robust forecasting in spatial autoregressive model with total variation regularization," Journal of Forecasting, John Wiley & Sons, Ltd., vol. 42(2), pages 195-211, March.
    16. José J. Quinlan & Fernando A. Quintana & Garritt L. Page, 2021. "On a class of repulsive mixture models," TEST: An Official Journal of the Spanish Society of Statistics and Operations Research, Springer;Sociedad de Estadística e Investigación Operativa, vol. 30(2), pages 445-461, June.
    17. Dawen Xia & Xiaonan Lu & Huaqing Li & Wendong Wang & Yantao Li & Zili Zhang, 2018. "A MapReduce-Based Parallel Frequent Pattern Growth Algorithm for Spatiotemporal Association Analysis of Mobile Trajectory Big Data," Complexity, Hindawi, vol. 2018, pages 1-16, January.
    18. Dong, C. & Li, S., 2021. "Specification Lasso and an Application in Financial Markets," Cambridge Working Papers in Economics 2139, Faculty of Economics, University of Cambridge.
    19. Lam, Clifford, 2008. "Estimation of large precision matrices through block penalization," LSE Research Online Documents on Economics 31543, London School of Economics and Political Science, LSE Library.
    20. Mohit Agrawal & Joseph G. Altonji & Richard K. Mansfield, 2019. "Quantifying Family, School, and Location Effects in the Presence of Complementarities and Sorting," Journal of Labor Economics, University of Chicago Press, vol. 37(S1), pages 11-83.

    More about this item

    NEP fields

    This paper has been announced in the following NEP Reports:

    Statistics

    Access and download statistics

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:arx:papers:2201.08326. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: arXiv administrators (email available below). General contact details of provider: http://arxiv.org/ .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.