【Title】Parameter Estimation in Epidemiology using Sparse Grid Interpolation
【Speaker】Dr. Nan Kong
【Host】Dr. Xiaolei Xie
【Time】14:30-15:30, January 15th, 2017, Monday
【Location】 Room N412, Shunde Building
【Abstract】We consider the problem of using time-series data to calibrate compartment-based epidemiological models. We propose a two-stage algorithm that identifies potentially optimal regions of the parameter space and directs computational effort towards resolving the dynamics and achieving good fitting in these regions. To facilitate this endeavor, we rely on sparse grid interpolation, a popular numerical discretization technique for the treatment of high dimensional, multivariate problems, to capture the dynamics underlying both global and local spaces. By employing cluster analysis techniques and metaheuristic algorithms, we show through two case studies that definitive gains in performance can be made to produce simulated outcomes consistent with available epidemiological data.
【Short Bio】Dr. Nan Kong is Associate Professor in the Weldon School of Biomedical Engineering at Purdue University. He is also a Faculty Advisor team member for the Purdue Regenstrief Center for Healthcare Engineering. Dr. Kong received his PhD in industrial engineering from the University of Pittsburgh in 2006. His research interest lies in population health management and policy, hospital and healthcare systems operations management, and big-data driven smart connected health operations research. He has published more than 40 peer-reviewed journal articles. He has received grants from NSF, NIH, AHRQ, and Air Force. Dr. Kong is an Associate Editor for the journal Institute of Industrial and Systems Engineers Transactions on Healthcare Systems Engineering. He has reviewed manuscripts for more than 20 journals, including Operations Research, Management Science, Mathematical Programming, Medical Decision Making, and American Journal of Public Health. He is the President-elect of the INFORMS Public Sector Operations Research section. He has organized tasks/clusters for several INFORMS meetings, including INFORMS 2016, 2017, INFORMS International 2012, 2015, and INFORMS Healthcare 2013, 2017.
 
All interested are welcome!