Ph.D. Profile: Angela Zhou

Angela Zhou is a doctoral student at Cornell’s School of Operations Research and Information Engineering (ORIE). She is based in New York City and works with her advisor, Assistant Professor Nathan Kallus, at the Cornell Tech campus on Roosevelt Island. Before studying at Cornell, Zhou earned her B.S.E. in Operations Research and Financial Engineering from Princeton University.

“In high school I was interested in a lot of subjects,” says Zhou. “I liked physics; I liked math; I liked anthropology.” Zhou attended a pre-engineering-themed high school and continued to read widely across the mathematical and social sciences. “Before I went to college I thought I was interested in physics and economics,” says Zhou. “As fields, they have very different ways of modeling and representing the world.”

Once Zhou got to Princeton she chose to major in Operations Research and Financial Engineering, thinking that it would be the closest approximation to applied math. She realized that operations research had long been studying how to model the world and capture insights from the complexities of real life. While studying at Princeton Zhou had several internships that showed her the broad relevance of these mathematical tools, from data science research to the tech industry.

When it came time to choose a graduate school, Cornell’s OR program was an excellent fit for Zhou. “My undergrad senior thesis was really inspired by work done by Cornell researchers,” says Zhou, “so I was familiar with the quality of their work. And when I looked at the Ph.D. program I appreciated the department’s blend  of theory and impactful practice.”

Zhou’s work with Kallus is most immediately focused on supporting better decision-making from observed data, in view of applications in healthcare and policy, using ideas from causal inference, machine learning, and optimization. But, in the end, Zhou says she wants to make tools that that will allow users from many domains to leverage their data. “I’m working on the evidence-based methodology used by practitioners from a variety of fields, and hoping to enhance its expressiveness and reliability to support better data-driven decisions. In the tech industry, we call it A/B testing; clinical trials in medicine and field experiments in the social sciences. In all of these areas, it can be expensive to run new experiments and thus we need to leverage observational data” says Zhou. “One of the reasons I was drawn to work with Nathan Kallus is my respect for some of his previous work on the opportunities and subtleties at the interface between data and decision-making.”

Zhou understands the importance of the human element in any decision-making tool she creates. “Behavioral economics reminds us that we need to consider how humans actually act on information, especially since they have richer contextual information, , when we think about how decisions are made in reality.” says Zhou. Zhou hopes that data-driven decision-making can empower practitioners and researchers to identify valuable insights from domain-specific data.