报告题目：Adaptive banding covariance estimation for high-dimensional multivariate longitudinal data

报 告 人：张伟平教授 中国科学技术大学

报告时间：2020年6月23日 上午 10:20-11:20

报告地点：腾讯会议

点击链接入会，或添加至会议列表：

https://meeting.tencent.com/s/YK0O6z7nCo4n

会议 ID：367 304 956

会议密码：200623

校内联系人：赵世舜 zhaoss@jlu.edu.cn

报告摘要：

Modeling the covariance matrix of multiple responses in longitudinal data plays a key role and is more challenging as compared to its univariate counterpart due to the presence of correlations among multiple responses. Using the modified Cholesky block decomposition, we impose an adaptive block banded structure on the Cholesky factor and sparsity on the innovation variance matrices via a novel convex hierarchical penalty and lasso penalty, respectively. The resulting adaptive block banding regularized estimator is fully data-driven and has more flexibility than regular banding estimators. An efficient alternative convex optimization algorithm is developed using ADMM algorithms. The resulting estimators are also shown to converge in an optimal rate of Frobenius norm, and row specific support recovery is established for the precision matrix. Simulations and real  data analysis show that the proposed estimator can be better able to reveal the banding sparsity pattern in the data.

报告人简介：

张伟平，中国科学技术大学教授，博导。主要从事纵向数据分析、风险理论、统计学习等领域中的统计理论和应用研究工作，先后在国内外学术期刊发表论文50余篇。主持了国家自然科学基金青年和面上项目、重点项目子课题等多个项目。曾获安徽省自然科学优秀论文一等奖、安徽省教学成果奖特等奖等。担任全国工业统计学教学研究会、中国商业统计学会、中国现场统计研究会环境与资源统计分会以及数据科学与人工智能分会等学会的理事。