I teach business and engineering students at the undergraduate, master's, doctoral, and executive levels. Here you'll find a summary of my classroom methods along with course information.
Quantitative analysis of a problem leads to insights and understanding that cannot be obtained through other means. Often, only those with a leaning toward mathematics have access to this kind of learning. But a quantitative education should be available to everybody. No matter a student's background, in my classroom complex concepts are made clear. Insight need not be masked by symbols and equations. It can be put into plain terms for all to appreciate.
Accessible quantitative education requires two things: stylized analysis and clear-cut delivery. Stylized analysis begins with a simpler version of a difficult problem, a scenario basic enough that it is amenable to examination, but with enough complexity that an optimal decision eludes intuition. Then, via a guided investigation, students discover the best way to manage various tradeoffs. Finally, the insight gleaned from the stylized analysis is scaled up to more complex settings, giving students a useful heuristic to make decisions in the real world.
Although stylized analysis begins with a simplified scenario, it may still seem daunting. To be accessible, quantitative instruction must be delivered in a way that makes equations less intimidating. This requires methods that distill the complex into something anyone can grasp. For example, clear-cut delivery may call for replacement of uncertainties and dynamics with easier-to-understand simulations. Random variables and calculus may be better explained via an intuitive Monte-Carlo-style exercise. Clear-cut delivery moves away from static symbols on a chalkboard toward a classroom that is visual and interactive.
Learning is a consequence of accessible analysis. When concepts that seem unreachable are brought within grasp, students have “Aha!” moments that stay with them for years. My teaching methods are designed to create these moments. Stylized analysis and clear-cut delivery tame what seems wild, helping students from all backgrounds obtain insight that can be had no other way.
This freshman seminar makes math personal. What should we do (or not) in a day or a lifetime? What amount of new and familiar is most fulfilling? How much mess is ok? This isn't a class for math majors. It's a course for everyone on thinking algorithmically, on learning about the fundamental structures of the problems we face, and ultimately on discovering something about ourselves.
I've always felt more comfortable with numbers and equations than with words and sentences. That's why I've spent my life studying them. Eventually, I realized that mathematics has just as much to say about living as philosophy or literature.
But you can't find this in textbooks or college coursework. Students dive into novels to see the world from a different perspective. They mine essays and poetry for insight. In contrast, the quantitative lens in their toolbox is nearsighted. We give them end-of-chapter problems and a few basic applications, but we leave the connection to their personal lives so distant that it's completely out of focus. I wanted to change that.
Algorithms to Live By is my effort to help students take a quantitative view on everyday problems as well as lifelong challenges, all in a way that's accessible to each student.
This doctoral-level course focuses on sequential decision making in the face of uncertainty. A variety of real-world challenges fall within this scope, including problems in business, engineering, the sciences, and health care.
In such problems, the decision maker is tasked with identifying alternatives that perform well not only now, but across some horizon. Because sequential decision problems cut across many domains, they are studied in various disciplines.
The course leverages advances in optimal control, Markov decision processes, and reinforcement learning to address modeling, policies, and dual bounds. It may be of particular interest to students who want to connect deterministic optimization techniques with stochastic solution strategies.
Saint Louis University doesn't maintain a doctoral program related to reinforcement learning. So I don't usually teach stochastic dynamic programming at home. However, I've enjoyed teaching these topics to students at other universities around the world.
How to match supply with demand at the firm level. Design, performance, and improvement of operations in manufacturing and service settings.
Undergraduate & Master's
How to match supply with demand across businesses. Managing the production, storage, and distribution of goods across a network of suppliers.
Master's
How to use predictive and prescriptive analytics to guide the allocation of limited resources. Quantitative methods to transform data into insight for better decision making.
Master's
How to hedge against the unknown by laying down resources across a range of outcomes. Why your forecast is wrong and what to do about it.
Executive
How to reduce, buffer, and pool variability. Operational strategies to hedge against fluctuations in supply and demand.
Executive