光华讲坛——社会名流与企业家论坛第6573期
主题:An introduction to high dimensional asymptotics 高维渐近导论(系列讲座)
主讲人:罗格斯大学 韩启阳副教授
主持人:西南财经大学统计学院 常晋源教授
时 间:2024年6月17日(周一)上午9:00-11:30 下午14:30-17:00
2024年6月18日(周二)上午9:00-11:30 下午14:30-17:00
2024年6月19日(周三)上午9:00-11:30 下午14:30-17:00
举办地点:西南财经大学光华校区光华楼10楼1003
主办单位:数据科学与商业智能联合实验室 统计学院 科研处
主讲人简介:
Qiyang Han is an Associate Professor of Statistics at Rutgers University. He received a Ph.D. in Statistics in 2018 from University of Washington under the supervision of Professor Jon A. Wellner. His research expands broadly in mathematical statistics and high dimensional probability, with a particular focus on empirical process theory and its applications to nonparametric and high dimensional statistics. He is a recipient of the NSF CAREER award in 2022, the Bernoulli Society New Researcher Award in 2023, and the David G.Kendall's Award in Mathematical Statistics in 2024.
韩启阳,罗格斯大学统计学副教授,于 2018 年获得华盛顿大学统计学博士学位,师从 Jon A. Wellner 教授。他的研究领域广泛,包括数理统计和高维概率,尤其侧重于经验过程理论及其在非参数和高维统计中的应用。他于 2022 年获得美国国家科学基金会 CAREER 奖,2023 年获得伯努利学会新研究员奖,2024 年获得 David G.Kendall 数理统计奖。
内容简介:
High dimensional asymptotics has emerged as a new theoretical paradigm to precise characterize the stochastic behavior of a large number of statistical estimators, finding a wide range of applications beyond the reach of standard theoretical methods. In these talks, we will briefly introduce three main theoretical approaches in this field. In the first part, we will discuss the leave-one-out method, originally introduced in the context of robust regression. In the second part, we will introduce a Gaussian process approach, currently known as the Convex Gaussian Min-Max Theorem framework. In the third part, we will discuss an algorithmic approach, known as the Approximate Message Passing method. We will provide both rigorous, theoretical foundations for these approaches, and illustrate the utility of these methods in some of the canonical statistical settings and the more recent interpolating estimators. Time permitting, I will also briefly discuss more recent theoretical developments in this field.
高维渐近理论已经成为一种新的理论范式,可精确描述大量统计估计量的随机行为,其应用范围超出了标准理论方法的范围。在本系列讲座中,我们将简要介绍该领域的三种主要理论方法。第一部分,我们将讨论留一法,该方法最初是在稳健回归的背景下引入的。第二部分,我们将介绍高斯过程法,目前称为凸高斯极大极小定理框架。第三部分,我们将讨论一种算法,即近似消息传递算法。我们将讲解这些方法严格的理论基础,并将在一些典型的统计设置和较新的插值估计量中说明这些方法的实用性。如果时间允许,我还将简要讨论该领域的更多最新理论进展。