报告人：方匡南教授 厦门大学
报告题目：Small Big Data: Integrative dimension reduction method for multiple data sets.
时 间：2017年5月5日 (星期五) 14:00 ~ 15:00
地 点：旗山校区理工北楼601报告厅
主 办：数学与计算机科学学院, 福建省分析数学及应用重点实验室, 数学研究中心
参加对象：相关专业教师和学生
报告摘要：For high dimension low sample size data, the analysis of single datasets often generates unsatisfactory results. The integrative analysis of multiple independent datasets provides an effective way of pooling information and outperforms single-dataset analysis and some alternative multi-datasets approaches including meta-analysis. Under certain scenarios, multiple datasets are expected to share common covariance matrices, that is, their principal component loadings have similarity in values,signs or sparsity structures. In this study, we consider sparse PCA in integrative analysis. We develop a penalization based approach, which is the first to take the similarity of principal component loadings into consideration. Theoretically it has the desired consistency properties.In simulation, it significantly outperforms the competing alternative methods.
专家简介：方匡南，厦门大学经济学院统计系教授、博士生导师，耶鲁大学博士后。入选福建省委组织部青年拔尖人才计划、福建省高校杰出青年科研人才培育计划、福建省高校新世纪优秀人才支持计划。曾先后发表论文70多篇，其中在 Journal of Multivariate Analysis、Nature子刊 Scientific Reports等国际权威期刊发表30多篇，在《管理科学学报》、《经济研究》、《统计研究》、《数量经济技术经济研究》等国内权威期刊发表40多篇。先后主持了国家自然科学基金面上项目、青年项目、国家社科基金重大项目子课题、国家统计局重大项目等多项。
