專題演講 主講人：Prof. Kuang-Yao Lee
(Department of Statistics, Operations, and Data Science, Fox School of Business, Temple University)
題 目：Graphical models: a review and recent developments
主講人：Prof. Kuang-Yao Lee
(Department of Statistics, Operations, and Data Science, Fox School of Business, Temple University)
時 間：112年12月29日(星期五)上午11:10-12:00
(上午9:50-10:10茶會於綜合一館428室舉行)
地 點：綜合一館427室
使用Google Meet線上直播，
演講開始前20分鐘可進入會議，請點選下列連結後按下「加入」即可
https://meet.google.com/pie-jmyd-cra
摘要
Graphical models (GM), a sub-field of machine learning, have played a fundamental role in a multitude of scientific disciplines, including epidemiology, genetics, sociology, and business, by facilitating the analysis of complex dependencies among random units. In this talk, we will delve into GM by first exploring some popular parametric approaches. Subsequently, we will introduce a novel nonparametric approach based on the concept of additive conditional independence (ACI)—originally proposed by Li et al. (2014, JASA). ACI revolutionizes the modeling of three-way relationships among random variables, aligning with the core principles of conditional independence but without the need for high-dimensional kernels smoothing. This unique characteristic strikes a balance between parametric and nonparametric models and hence enhances scalability, particularly for handling large graphs. To cap off the talk, we will leverage ACI to present a comprehensive GM framework, applicable in both parametric and nonparametric contexts, and capable of accommodating random variables or random functions as nodes.