报告名称:VC dimension and related problems
报 告 人:李本崇 副教授
报告时间:2021年10月9日 9:00-12:00
报告地点:B5-201
报告人介绍:
李本崇,西安电子科技大学华山菁英副教授,统计学和应用统计方向硕士生导师。2001--2012年在东北师范大学读书,获统计学博士学位。主要从事图模型和代数统计领域的研究,在国际知名统计学期刊 Pattern Recognition,Statistica Sinica,中国科学-数学等期刊发表论文 12 篇(一作8篇,二作1篇,三作3篇)。曾主持一项国家自然科学基金青年基金,一项陕西省自然科学基金青年基金;现主持一项国家自然科学基金面上项目,一项陕西省自然科学基金面上项目。中国现场统计研究会大数据统计分会、数据科学与人工智能分会理事;全国工业统计学教学研究会、青年统计学家协会理事、数字经济与区块链技术分会理事(副秘书长);陕西省数学会统计分会理事。
报告简介:
To characterize the family of learnable classes in the setup of binary valued classification with the zero-one loss, Vladimir Vapnik and Alexey Chervonenkis coined the combinatorial measure called the Vapnik-Chervonenkis dimension (VC dimension) in 1970. In this talk, I will introduce the concept of VC dimension and show several examples as well as the fundamental theorem of statistical learning. Then I shall present our work in this field, we show that the value of VC dimension of the concept class induced by a discrete undirected graphical model equals dimension of the toric ideal corresponding to this undirected graphical model as long as the toric ideal is nontrivial. Finally, I state a few open problems related to VC dimension.