SOCR News ISS JSM 2025
Contents
SOCR News & Events: 2025 Joint Statistical Meeting, Nashville, TN
The 2025 Joint Statistical Meeting (JSM) will take place August 2-7, 2025 in Nashville, TN. The annual event will feature an invited special session entitled Statistical Inference and AI Modeling of High-Dimensional Longitudinal Data.
Session Logistics
- Title: Statistical Inference and AI Modeling of High-Dimensional Longitudinal Data
- Organizer: Ivo Dinov (Michigan)
- Chair: Chunming Zhang, University of Wisconsin-Madison
- Speakers: Changbo Zhu, University of Notre Dame, Yi Zhao, Indiana University, and Yueyang Shen (Michigan)
- Discussant: Ivo Dinov (Michigan), Statistics Online Computational Resource (SOCR)
- Date/Time: ...tbd... US ET
- Venue: ...tbd...
- Registration: ...tbd...
- Conference: ...tbd...
- Format: ...tbd...
Sponsors
- Section on Statistics in Imaging
- International Association for Statistical Computing
- International Statistical Institute
Session Description
In support of the for JSM 2025 theme, "Statistics, Data Science, and Al Enriching Society," this invited special session will bring together a diverse group of academic researchers, each contributing to the cutting-edge intersections of statistical learning methods, Al applications, and high-dimensional data analysis.
- Changbo Zhu (Notre Dame) will delve into Testing independence for sparse longitudinal data exploring the fundamental question of independence of a pair of random functions. This problem becomes quite challenging when the random trajectories are sampled irregularly and sparsely for each subject. In other words, each random function is only sampled at a few time-points, and these time-points vary with subjects. Furthermore, the observed data may contain noise. To the best of our knowledge, there exists no consistent test in the literature to test the independence of sparsely observed functional data. We show in this work that testing pointwise independence simultaneously is feasible. The test statistics are constructed by integrating pointwise distance covariances and are shown to converge, at a certain rate, to their corresponding population counterparts, which characterize the simultaneous pointwise independence of two random functions. The performance of the proposed methods is further verified by Monte Carlo simulations and analysis of real data..
- Yi Zhao (Indiana University) will present on "Longitudinal and Covariance Regression of High-Dimensional Data," addressing the challenges and methodologies for analyzing data with both temporal and complex covariate structures.
- Yueyang Shen (University of Michigan) will then discuss "Deep Learning Invariance and Statistical Estimation Equivariance with Applications," offering insights into how deep learning models can maintain robustness and consistency across varying conditions, and how this is applied in real-world data scenarios.
In this session, Changbo Zhu, Yi Zhao, and Yueyang Shen will present cutting-edge methodologies that harness the power of statistical learning, topological modeling, and deep learning to advance the analysis of complex, high-dimensional spatiotemporal data in various scientific domains. Speakers will demonstrate topological methods for handling challenges related to neuroimaging data complexity and scale and provide new perspectives on brain connectivity and function. Recent work on longitudinal and covariance regression in high-dimensional data will expose temporal structures involving a large number of covariates. The topics will cover deep learning invariance and equivariance, which characterize the robustness and generalizability of various AI models in real-world applications. All talks will emphasize challenges and opportunities in statistical and Al techniques to extract meaningful unbiased insights from complex data. This session aligns with the overarching conference theme "Statistics, Data Science, and Al Enriching Society" and will highlight applications in neuroscience, longitudinal studies, and AI forecasting, and trustworthy decision-making.
Translate this page:
