题目：Generalized unbiased estimating equation in penalized parametric quantile regression for life expectancy determinants

汇报人： 虞克明

会议时间：2025年6月23日(周一)  14：30-15：30

地点： 综合楼615会议室

报告人简介：虞克明，英国伦敦布鲁内尔大学统计与数据科学讲习教授(Chair Professor)、 数学学科研究影响中心主任；英国皇家统计学会会士、英国社科基金 (ESRC) 评审专家成员、英国自科基金 (EPSRC)评审专家成员 、欧洲科学基金(ESF) 评审专家成员。目前是《Journal of the Royal Statistical Society-C》副主编，曾担任过《Journal of the American Statistical Association》、《Journal of the Royal Statistical Society-A》等多家统计权威期刊的副主编。目前他主要从事回归分析、 非参数统计、 机器学习、 贝叶斯推断、大数据及非常小的数据分析等方面的理论和方法研究，是贝叶斯分位数回归方法的开拓者，先后在JASA、JRSSB、JRSSA、JRSSC、JOE、JBES、Bernoulli等统计学顶级刊物上发表论文150多篇。

摘要：

Quantile regression for longitudinal data has received considerable attention over the past decade. Recent research has primarily focused on developing unbiased estimating equations or parametric modeling of quantile regression coefficients. The former is essential for the accurate, valid, and efficient estimation of model parameters, while the latter enhances flexibility, interpretability, and the capacity to capture the dynamic effects of covariates across quantile levels.  However, no existing method effectively incorporates both.  In this paper, we propose a method that integrates both features, referred to as the generalized unbiased estimating equation approach to penalized parametric quantile regression for longitudinal data.

The model is based on a penalized longitudinal parametric quantile regression, incorporating a novel smoothing estimator. Notably, our method outperforms existing approaches, offering improved estimation efficiency even when working correlation structures are misspecified. The proposed estimators are asymptotically efficient under regularity conditions and are easy to implement. We evaluate the proposed method through simulations and apply it to investigate the key determinants of global life expectancy, considering factors related to geography, demographics, economics, environment, infrastructure, healthcare, and lifestyle. The findings provide valuable insights for national policies aimed at improving life expectancy.