报告人： Weiwei Zhuang（University of Science and Technology of China）
时间：2026-09-10 13:30-14:50
地点：Room 217, Guanghua Building 2
Abstract:
We develop a computationally efficient and robust estimation procedure for finite Gaussian mixtures based on a Gaussian-tapered Hermite representation. The tapering induces a transformed mixing distribution that preserves the algebraic structure underlying moment recovery while substantially reducing the influence of extreme observations. This representation leads to the robust denoised method of moments (RDMM), which combines moment projection with an explicit recovery of the original mixing distribution. When the structural variance is unknown, we further develop an estimator based on the positive-semidefinite geometry of Hankel moment matrices. We establish the tapered moment identity, derive the associated recovery procedures, and show that the resulting estimator has a bounded influence function. Simulations show substantial gains in robustness and numerical reliability under contamination. Notably, these gains do not come at the expense of efficiency under the uncontaminated model; on the contrary, RDMM improves upon DMM across the configurations considered.
About the Speaker:
Weiwei Zhuang is an Associate Professor and doctoral supervisor at the University of Science and Technology of China (USTC). She is a recipient of the Anhui Provincial Distinguished Young Scholars Fund. Her research interests include stochastic comparisons, statistical testing, and risk management.
She has published numerous papers in internationally recognized journals, including Biometrika, Insurance: Mathematics and Economics, Journal of Multivariate Analysis, and Statistica Sinica. Her research has also been featured in scholarly monographs published by Elsevier and Springer. She has served as principal investigator for four projects funded by the National Natural Science Foundation of China, including three General Program projects and one Young Scientists Fund project.
Dr. Zhuang has received the First Prize of the Anhui Provincial Natural Science Paper Award and has been recognized as an Outstanding Young Teacher of Anhui Province. As a key team member, she has also contributed to a National First-Class Undergraduate Course and to teaching achievements recognized with top provincial awards.
Your participation is warmly welcomed!
欢迎扫码关注北大统计科学中心公众号，了解更多讲座信息!
