“数理讲堂”2024年第8期:Subsampling for Big Data Expectile Regression with Errors-in-Variables

发布时间:2024-04-26 供稿:数理学院 分享至:

主题:Subsampling for Big Data Expectile Regression with Errors-in-Variables

时间:2024年04月26日 15:00-16:30

地点:15号楼518会议室

主持人:姜荣 教授

报告人简介:

王明秋,曲阜师范大学统计与数据科学学院,博士,教授,博士生导师。主要研究方向高维数据分析,大数据子抽样。中国现场统计研究会统计调查分会常务理事、山东省应用统计学会常务理事、试验设计分会理事、数据科学与人工智能分会理事。先后主持国家自然科学基金面上项目、青年基金和多项省部级基金。

讲座简介:

Subsampling is a technique for extracting subdatasets from a complete dataset and estimating parameters of interest based on the selected subdata. However, existing literature assumes that covariates can be directly and accurately observed. To address the challenges posed by measurement errors and heterogeneity in big data, this study considers the optimal subsampling algorithm based on expectile regression. Additionally, we provide asymptotic properties for the subsampling estimator. Through simulation studies and real data analysis, we demonstrate the effectiveness of our proposed method compared to other algorithms that do not account for measurement errors.

打印
上一篇:下一篇: