Join Zoom Meeting
Kimia Ghobadi, PhD
John C. Malone Assistant Professor of Civil and Systems Engineering
Center for Systems Science and Engineering
Johns Hopkins University
Abstract: Many applicationsutilize optimization models, but the correct parameters for the optimizationmodels are usually hard to know. While the process of optimal decision-makingmay be unknown, there are often solutions (or observations) of the system thatare available. In this talk, we focus on Inverse Optimization techniques toinfer the parameters of optimization models from a set of observations. Inverseoptimization can be employed to infer the utility function of a decision-makeror to inform the guidelines for a complex process. We present a data-driveninverse linear optimization framework (called Inverse Learning) that finds theoptimal solution to the forward problem based on the observed data. We discusshow combining inverse optimization with machine learning techniques can utilizethe strengthsof both approaches. Finally, we validate the methods usingexamples in thecontext of precision nutrition and personalized daily dietrecommendations.
Bio: Kimia Ghobadi is aJohn C. Malone AssistantProfessor of Civil and Systems Engineering and amember of the Malone Center for Engineering in Healthcare, the Center forSystems Science and Engineering (CSSE), and the Center for Data Science inEmergency Medicine. Prior to joining JHU, she was a postdoctoral fellow at MITSloan School of Management and obtained her PhD from the University of Torontoin Industrial Engineering. Her research interests are in developing inverse andforward optimization models, real-time algorithms, and analytics technics withapplicationin healthcare systems including healthcare operations and medicaldecision-making.