Nonparametric variable selection with penalized neural network

讲座通知

 

2016年12月22日(星期四)

下午 3:30

 

信息管理与工程学院308会议室

上海财经大学(第三教学楼西侧)

上海市杨浦区武东路100号

主题

Nonparametric variable selection with penalized neural network

(深度学习中的变量选取问题)


主讲人

Ye Luo

University of Florida

Education

PHD- Massachusetts Institute of Technology, 2015

BS- Mathematics, Economics, Massachusetts Institute of Technology, 2010


Research Interests

Econometrics, Finance, Industrial Organization


讲座简介

Neural Network has been popular amongst machine learning in recent years. We study high-dimensional non-parametric variable selection problem using neural network. We derived a data-driven penalized method to select input variables as well as the number of nodes in the hidden layers.