Academic Report of School of Mathematical Sciences [2025] No. 132

(Series Report for High-Level University Construction No. 1234)

Title: Penalized Weighted Generalized Estimation Equations for High-Dimensional Longitudinal Data with Informative Cluster Size

Speaker: Researcher Jiang Xuejun (Southern University of Science and Technology)

Time: 15:00–16:00, November 30, 2025 (Sunday)

Location: Classroom 1, Huixing Building, Yuehai Campus, Shenzhen University

Abstract:

High-dimensional longitudinal data have become increasingly common in recent research. Penalized Generalized Estimating Equations (GEE) are often used to model such data. However, when the outcome variable under study is correlated with the cluster size, the desirable properties of the GEE approach may be compromised, a phenomenon known as "informative cluster size." This paper explicitly addresses the impact of informative cluster size and proposes a novel weighted GEE method to mitigate its effects, while extending the penalized approach to high-dimensional settings. We demonstrate that the penalized weighted GEE method maintains consistency in both model selection and estimation. Theoretically, we prove that, assuming the true model is known, the proposed penalized weighted GEE estimator is asymptotically equivalent to the Oracle estimator. This result indicates that the penalized weighted GEE method preserves the excellent properties of the GEE approach and exhibits strong robustness to informative cluster size, thereby extending its applicability to highly complex scenarios. Validation through simulations and real-data applications further demonstrates that the penalized weighted GEE method outperforms existing alternative approaches.

Speaker Biography:

Jiang Xuejun is the Associate Head (Teaching and Undergraduate Affairs) of the Department of Statistics and Data Science, a Researcher, and a Doctoral Supervisor at the Southern University of Science and Technology (SUSTech). He received his Ph.D. from the Department of Statistics at The Chinese University of Hong Kong in 2009 and conducted postdoctoral research there from 2009 to 2010. He joined SUSTech in July 2013 as an Associate Professor. He was selected for the Shenzhen Peacock Plan for Overseas High-Level Talents (2016) and was named an Outstanding Teacher of Shenzhen (2018). He has presided over and completed more than 10 projects, including those from the National Natural Science Foundation of China, the Natural Science Foundation of Guangdong Province, and the Shenzhen Basic Research General Program. His research directions and interests involve statistical analysis of large-scale complex data, high-dimensional statistical inference, financial/applied statistics, transfer learning, representation learning, and auxiliary learning. He has published nearly 60 SCI/SSCI papers in top-tier statistics journals such as Biometrika, Bernoulli, Statistica Sinica, Statistics and Computing, The Econometrics Journal, Financial Innovation, Science China Mathematics, and Scientia Sinica Mathematica, along with 2 authorized patents and one English textbook publication.

All faculty and students are welcome!

Host: Hu Xianghong

School of Mathematical Sciences

November 21, 2025