报告题目：Distributed Censored Quantile Regression: Convolution Smoothing and Communication Efficiency

报告时间：2026年7月21日上午10:00

报告地点：南湖校区综合三报告厅

主办单位：伟德体育

报告人：马慧娟

报告人简介：马慧娟，华东师范大学统计伟德体育与统计交叉科学研究院副教授。在统计学期刊 Journal of the American Statistical Association (JASA), Biometrika, Biometrics, Journal of Business & Economic Statistics (JBES), Statistica Sinica等期刊发表论文二十余篇。先后主持上海市浦江人才项目，国家自然科学基金青年项目、面上项目和重点项目子课题等。参与教育部学科突破先导项目、国家自然科学基金重点项目、科技部重点研发项目和上海市科委重点项目等。

摘要：Censored quantle regression (CQR) has become a popular framework for analyzing survival outcomes, yet conventional estimation procedures facecomputational challenge when scaled to massive datasets. In such modern large-scale settings, individual-level survival data cannot be freelyaccessed or centrally pooled due to privacy concems and storage limitations, rendering distributed computation increasingly indispensable. The nondifferentiability of the CQR loss further complicates optimzation, making exsting methods computationaly expensive and poorly suited fordistributed data structures. To overcome these challenges, we propose a communication-efficient distributed CQR framework that leveragesonvolution smoothing to construct a globaly smooth and convex objective, thereby enabling quasi-Newton optimization in large- scale distributednvironments. The proposed approach effectively balances computational scalability and communication efficiency. We establish the Bahadurepresentation and asymptotic properties of the convolution-smoothed CQR estimator on a single machine, and further derive the convergence rateof the distributed quasi-Newton updates. Extensive simulations and real-world applications demonstrate that the proposed method achievesstatistical acuracy comparable to single- machine ful- sample algorithms, while outperforming existing distributed CQR methods in bothcomputational and communication efficiency.