讲座题目：Stochastic Gradient Estimation:Efficient Learning Technique Under Uncertainty

主讲人：彭一杰助理教授，北京大学工业工程与管理系 

时间： 2019年6月29日（周六） 

地点：文科楼管理学院资料室2423A 15:00-16:00 

主持人：马利军教授 

内容简介：

 Stochastic gradient estimation plays a central role in sensitivity analysis andgradient-based optimization. Particularly, it serves as the most important toolfor training artificial neural networks in deep learning. We propose ageneralized likelihood ratio (GLR) estimator in a framework that can handlediscontinuous sample performances with structural parameters. This work extendsclassic unbiased stochastic derivative estimators to a setting where they didnot previously apply. The GLR estimator can deal with many applications in asingle umbrella and preserves the single-run efficiency. Applications of GLR onfinancial engineering, risk management, and artificial intelligence will bepresented.

主讲人简介：

Dr. Yijie Pengis currently an assistant professor of the Department of Industrial Engineeringand Management at Peking University (PKU). He received his Ph.D. from theDepartment of Management Science at Fudan University and his B.S. degree fromthe School of Mathematics at Wuhan University. Before joining PKU, he worked asan assistant professor at George Mason University, and postdoctoral scholar atFudan University and R.H. Smith School of Business at University of Maryland atCollege Park. Many of his publications appear in high-quality journalsincluding Operations Research, IEEE Transactions on Automatic Control, INFORMSJournal on Computing, Journal of Discrete Event Dynamic System, andQuantitative Finance. His research interests include stochastic modeling andanalysis, simulation optimization, machine learning, data analytics, andhealthcare.

 Welcome to join this lecture!

深圳大学管理学院管理科学系