题目:Dynamic Covariance Models
时间:2015年6月9日(周二)15:30-16:00
地点:6号楼415教室
主讲人:谌自奇博士
主办单位:数学与统计学院
Abstract:
An important problem in contemporary statistics is to understand the relations among p variables,usually with p much larger than the sample sizen. Recent efforts have been focusing on modeling static covariance matrices where pairwise covariances areconsidered invariant. In many real systems, however, these pairwise relations often change. To characterize the changing correlations in a high dimensional system, we study a class of dynamic covariance models (DCMs) assumed to be sparse, and investigate for the first time a unified theory for understanding their non-asymptotic error rates and model selection properties. In particular, in the challenging high dimension regime, we highlight a new uniform consistency theory in which the sample size can be seen as $n^{4/5}$ when the bandwidth parameter is chosen as for accounting for the dynamics, and this result holds uniformly over a range of the variable used for modeling the dynamics. The convergence rate bears the mark of the familiar bias-variance trade-off in the kernel smoothing literature. We illustrate the results with simulations and the analysis of a neuroimaging dataset.
主讲人简介
教育背景
2006.9-2012.6,东北师范大学数学与统计学院,概率论与数理统计专业,硕博连读。
2002.9-2006.6,湖南师范大学数学与计算机科学学院,数学与应用数学专业,本科阶段的学习。
学术服务
Referee for Statistica Sinica;Comput.Statist. Data Anal.
研究兴趣
纵向数据分析,剖面似然研究,高维数据充分降维。
项目资助
国家自然科学基金青年基金,11401593,基于剖面似然的统计推断,22万元,在研,主持。
教育部高等学校博士点基金,20130162120086,高维协方差矩阵的非参数模型及其基于阈值规则的估计,4万元,在研,主持。
第七批中国博士后科学基金特别资助,2014T70778,高维动态图模型研究,15万元,在研,主持。
第53批中国博士后科学基金,高维稀疏非参数协方差矩阵估计及其应用,2013M531796,5万元,在研,主持。
国家自然科学基金青年基金,11301245,有序约束下动态处理效应的统计方法研究,22万元,在研,第一合作者。
国家自然科学基金面上项目,我校有序数据的统计推断,26万元,已结题,参加。
高等院校博士生创新基金,纵向数据中半参数模型的有效估计方法与应用,09SSXT116,2万元,已结项,主持。