当前位置: 首 页 - 科学研究 - 学术报告 - 正文

bat365在线平台、所2019年系列学术活动(第79场): 林金官教授 南京审计大学

发表于: 2019-05-30   点击: 

报告题目:Estimation and test of jump discontinuities in varyingcoefficient models with empirical applications

报 告 人:林金官教授 南京审计大学统计科学与大数据研究院

报告时间:2019639:00-9:40

报告地点:数学楼第二报告厅

报告摘要:

Varying coefficient models are very important tools to explore the hidden structurebetween the response and its predictors. This paper focuses on estimatingand diagnosing jump discontinuities in coefficient functions. A nonparametricprocedure is proposed to estimate jump discontinuities based on the Nadaraya-Watson kernel smoothing and least-squares fitting, and asymptotic propertiesof resulting estimators are derived. Then, a jump size-based test statistic isdeveloped for checking whether the estimated jump discontinuities are true. Acomputationally feasible approximation is derived for critical values of its limitingnull distribution. Monte Carlo simulations are conducted to assess the finitesample performance of the proposed methodologies, and an empirical example is discussed.

报告人简介:

   林金官,男,1964年生,博士、教授。现任南京审计大学统计科学与大数据研究院经理,兼任中国现场统计研究会工程概率统计学会副会长、中国现场统计研究会资源与环境统计学会副会长、教育部统计学类教学指导委员会委员、中文核心期刊《系统科学与数学》与《数理统计与管理》杂志编委等。