Dear colleagues,
The IISE Quality Control and Reliability Engineering (QCRE) Division and the Data Analytics and Information Systems (DAIS) Division will jointly organize a webinar series in the coming weeks. The talk, titled " Sparse Regression Analysis of Mixed Multi-Responses," will be given by Dr. Xiaoyu Chen on Monday, April 8th, from 1:00 PM – 2:00 PM EDT.
Dr. Xiaoyu Chen is an Assistant Professor of Industrial and Systems Engineering at the University at Buffalo. He received his Ph.D. from the Grado Department of Industrial and Systems Engineering and earned his M.Eng. degree in Computer Science from Virginia Tech in 2021. His research focused on statistical learning and machine learning with applications to cybermanufacturing systems and perioperative medicine. Dr. Chen is a member of the Institute for Operations Research and the Management Sciences (INFORMS), the Institute of Industrial and Systems Engineering (IISE), and the Association for Computing Machinery (ACM). His ongoing research projects are funded by the American Heart Association, America Makes, state government agency, and industry partners.
Abstract of this Webinar:
Performance metrics of real-world systems are usually defined with correlated but mixed multi-responses. For example, the runtime performance metrics of a Fog computing network can easily include both continuous, binary, and counting variables, which follow different types of distributions. Such mixed multi-responses have posed significant challenges to existing regression models. Current practice suggests one analyzing mixed responses separately by using different models, which loses the information contained in the hidden association among mixed responses. Intuitively, jointly modeling mixed multi-responses is more desirable since quantifying hidden association can further enhance the predictive power by imposing regulation among responses. However, how to define correlation/covariance among mixed multi-responses remains an open question, which prevents the direct adoption of multivariate generalized linear regression model. Our objective is hence to provide a sparse regression model that can (1) quantify the hidden association among mixed multi-responses, (2) select important variables and variable groups, and (3) quantify the uncertainty of both model coefficients and the hidden association. In this talk, I will present studies on (1) a multivariate regression of mixed responses (MRMR) model with application to a visualization evaluation problem; and (2) a Bayesian MRMR model that considers both individual and group variable selection with application to a runtime performance metrics prediction problem in Fog computing network. In addition, future research opportunities will be discussed.
To register for this event, please visit: https://events.teams.microsoft.com/event/07fbe7b7-6ade-4127-8a8b-9fb7ab65f5ad@1bf21f59-ba3f-4757-8a65-44b4acb89f42 (The flyer is also attached with this email)
More webinars and recordings can be found at: https://www.iise.org/details.aspx?id=643
For more information about this webinar, please feel free to contact the event organizers:
Yu (Chelsea) Jin: yjin@binghamton.edu
Xiaowei Yue: yuex@tsinghua.edu.cn
Syed Hasib Akhter Faruqui: shf006@shsu.edu
You are welcome to redistribute this announcement to your networks.
Sincerely,
IISE QCRE & DAIS Divisions