Identifying nonlinear variation patterns with deep autoencoders. (2nd December 2018)

• Record Type: Journal Article
• Authors: Howard, Phillip; Apley, Daniel W.; Runger, George
• Is Part Of: IISE transactions. Volume 50:Number 12(2018)
• Page Start: 1089
• Page End: 1103
• Publication Date: 2018-12-02
• Subjects: Profile data -- variation pattern -- autoencoder -- autoassociative neural network -- deep learning -- visualization; Industrial engineering -- Periodicals; Systems engineering -- Periodicals; Industrial engineering; Systems engineering; Electronic journals; Periodicals; 670.285
• DOI: 10.1080/24725854.2018.1472407
• Languages: English
• ISSNs: 2472-5854

Abstract

Abstract: The discovery of nonlinear variation patterns in high-dimensional profile data is an important task in many quality control and manufacturing settings. We present an automated method for discovering nonlinear variation patterns using deep autoencoders. The approach provides a functional mapping from a low-dimensional representation to the original spatially-dense feature space of the profile data that is both interpretable and efficient with respect to preserving information. We compare our deep autoencoder approach to several other methods for discovering variation patterns in profile data. Our results indicate that deep autoencoders consistently outperform the alternative approaches in reproducing the original profiles from the learned variation sources.