In my research group, we’re tackling some fascinating problems in areas like portfolio optimization, revenue management, assortment optimization, and online advertising, among others. These fields often require pulling together insights from seemingly unrelated disciplines, and that’s where things get really interesting. For example, in the context of assortment optimization, one of the models we studied with collaborators—which has since become a standard in the field—was based on a product substitution approach using a Markov chain model. What was surprising, and exciting, was that this assortment optimization problem turned out to be equivalent to an optimal stopping problem! This insight allowed us to design a highly efficient method for solving unconstrained assortment optimization problems, a key advance for retailers and marketers trying to maximize their offerings.
Beyond this, we often find ourselves connecting ideas from operations research, economics, and machine learning to address these complex optimization problems. Whether it’s optimizing portfolios under uncertainty or managing revenues in dynamic markets, our work spans a broad range of applications. Each of these areas presents its own unique challenges, and we’ve been developing new models and algorithms that are both innovative and practical. The ability to cross disciplines and bring together different methodologies is a core strength of our group, and it’s leading to some really exciting results.