Skip to content
Thiago Serra
Thiago Serra

Thiago Serra

Assistant Professor of Business Analytics
University of Iowa
Bio

Thiago Serra recently joined the University of Iowa’s Tippie College Business as Assistant Professor of Business Analytics, following 5 years as Assistant Professor of Analytics and Operations Management at Bucknell University’s Freeman College of Management. Previously, he was Visiting Research Scientist at Mitsubishi Electric Research Labs from 2018 to 2019, and Operations Research Analyst at Petrobras from 2009 to 2013. He has a Ph.D. in Operations Research from Carnegie Mellon University’s Tepper School of Business, for which he received the Gerald L. Thompson Doctoral Dissertation Award in Management Science in 2018. He obtained his pre-doctoral training in Brazil at the University of Campinas (Unicamp) and at the University of Sao Paulo (USP). During his Ph.D., he was also awarded the INFORMS Judith Liebman Award. His research at the intersection of machine learning and mathematical optimization is supported by the National Science Foundation (NSF).


Constraint Learning in Practice: What is Machine Learning Doing Inside My Optimization Model, and How Can My Optimization Solver Deal with It?

Constraint learning is a new paradigm of modeling optimization problems with the help of machine learning. For example, we can learn missing constraints from available data, or we can avoid writing a nonlinear objective function by learning a simpler replacement for it. We usually achieve that by using a neural network as part of the optimization model. However, the devil is in the details: making the wrong choices for how you build this neural network may lead to an optimization model that does not work at all. In the same way that there are good practices in how you should formulate a classic optimization model (without machine learning in it), there are also good practices in how you should formulate constraint learning models. In this talk you will learn why that happens, and how to design better constraint learning models.

Essential / Professional