English Version

首页» 学术科研» 讲座信息

[2017/7/13]Modeling of Spatio-Temporal Data and Some Applications




Modeling of Spatio-Temporal Data and Some Applications


报告人 Dr. Xiao Liu

Department of Industrial Engineering

University of Arkansas


主持人: 张玺 特聘研究员





Under the classic geo-statistical modeling paradigm, the modeling of spatio-temporal data often requires the full specification of the joint space-time covariance structure. However, for many real-life spatio-temporal processes (such as urban air pollution, extreme precipitation process, surface quality degradation, etc.), the space-time covariance structure is so complex that can hardly be specified or even envisioned. On the other hand, underlying physics (if available) often provides valuable insights into the statistical modeling and can be used to motivate a new class of statistical modeling approaches. In this talk, two real application examples, including the modeling of urban air quality data and weather radar image data, are presented to illustrate the idea of physics-based modeling of spatio-temporal data. In the first example, a spatio-temporal air quality model is proposed motivated from a widely used scalar transport equation describing the dispersion of air-borne pollutants, while in the second example, a space-time conditional autoregressive model is presented based on the classic forced-advection problem under a Lagragian type of advection scheme. Some on-going and future research topics are also discussed.



Dr. Xiao Liu is an Assistant Professor at the Department of Industrial Engineering, University of Arkansas. Before that, he was a Research Scientist at IBM Thomas J. Watson Research Center, New York (2015?2017) and IBM Research Collaboratory Singapore (2012?2015), and served as an Adjunct Assistant Professor at the ISE Department, National University of Singapore (2013?2016). His research focuses on engineering probability and statistics, spatio-temporal modeling, big data analytics, and various data-driven methodologies in broad areas such as quality and reliability, manufacturing yield prediction, preventive maintenance, urban air quality modeling, extreme weather events prediction, etc. He has published in peer-reviewed journals including Technometrics, the Annals of Applied Statistics, JQT, IIE Transactions, IEEE Transactions on Reliability, etc. He received the IBM Outstanding Technical Achievement Award in both 2015 and 2017, the prestigious Ralph A. Evans/P.K. McElroy Award at the 2011 Reliability and Maintainability Symposium (RAMS), and the best referred paper award from the QSR section at INFORMS 2016. He is on the editorial board of Quality and Reliability Engineering International.