Online Program Home
My Program

Abstract Details

Activity Number: 177 - Section on Statistical Learning and Data Science CPapers 2
Type: Contributed
Date/Time: Monday, July 30, 2018 : 10:30 AM to 12:20 PM
Sponsor: Section on Statistical Learning and Data Science
Abstract #329146
Title: Sufficient Dimension Reduction Using Deep Neural Networks
Author(s): Yixi Xu* and Xin Zhang and Xiao Wang
Companies: Purdue University and Florida State University and Purdue University
Keywords: neural networks; dimension reduction

Deep neural networks are increasingly favored due to their unprecedented success in a variety of classification and regression tasks on large datasets. In this paper, we develop an efficient algorithm for estimate the central mean subspace (CMS) based on the deep neural network. The advantage of our approach is that the dimension reduction and prediction (or estimation of the conditional mean function) are simultaneously achieved. In addition, theoretical analysis establishes both consistency of the estimation of CMS and the generalization error bound for prediction. The numerical results indicate that the proposed method can work very well and can achieve the state-of-the-art performance for benchmark datasets.?

Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program