【经管大讲堂2018第108期】

作者:时间:2018-12-05浏览:213供图:审阅:来源:南京航空航天大学

字体:

Functional Linear Regression for Mixed Tensorial and Functional Predictors

Jionghua (Judy) Jin (金炯华)

 Professor

Industrial & Operations Engineering Department, University of Michigan

Abstract:

Time:2018年12月26日10:00

Add:将军路校区经管楼702室

With the rapid development of advanced sensing techniques, various types of system operational signals can be feasibly collected. Understanding the relationship between the system performance and the associated operating signals is one of the most important tasks for ensuring the system performance and making effective decisions for smart manufacturing. This talk will present a general regression model for a scalar system response with mixed types of predictors including images, functional profile signals, and scalar attributes. To represent a set of time-dependent images, a third order tensor is employed for preserving not only the spatial correlation within one image but also the temporal dependency among multiple images. The advantage of the presented model can handle both tensorial and functional predictors without performing stack-up on tensorial predictors. To estimate the model, a new algorithm is developed by iteratively running three sub-problems for estimating each of tensorial, functional, and scalar predictors. A systematic approach is proposed to automatically determine the running sequence among three sub-problems based on the contribution of each type of predictors to the system response. The performance of the proposed model is evaluated using simulations and a real-world case study.

--------------------------------------------------------------------------------------------------------

Short Bio for Professor Jionghua (Judy) Jin

Jionghua (Judy) Jin is currently a professor in the Department of Industrial and Operations Engineering and the Director of Manufacturing Program at the University of Michigan. She received her PhD in Industrial and Operations Engineering at the University of Michigan in 1999.

Dr. Jin’s research focuses on developing new data fusion methodologies with broad applications in both manufacturing and service industries.  She has received numerous awards the NSF CAREER Award, and the prestigious Presidential (PECASE) Award, 12 Best Paper Awards since 2005 etc.  She is currently the Editor for IIE Transactions on Quality and Reliability Engineering, and was Vice President of INFORMS and the President of QCRE division in IIE. She is a Fellow of IIE, a Fellow of ASME, an elected senior member of ISI, a senior member of ASQ, and a member of IEEE, INFORMS, and SME.

More information about Dr. Judy Jin can be found at http://jhjin.engin.umich.edu/