Before joining Columbia Business School’s PhD program, Tianyi “Tony” Zhang frequently jumped between different areas of study. “During my undergraduate year, I focused on systems biology and then non-equilibrium statistical physics. For my master’s, I was fascinated by topological data analysis and discrete differential geometry,” he says. “I transferred from area to area because I wanted to explore and learn as many different topics as possible from a research standpoint.”
Zhang eventually chose to focus his research on the intersection of data science and operations research, because of the opportunities to make a social and scientific impact on society. “Specifically, my PhD area of study lies in reinforcement learning, applied probability, sequential decision-making, and robust machine learning” Zhang says. “My research focus areas include: 1) designing algorithms with guarantees on statistical and computational performance especially in terms of sample complexity and robustness, and 2) deciphering the intrinsic hardness of specific learning tasks.”
He chose CBS’s DRO program because of its reputation and world-class faculty. “When I heard Professor Daniel Russo and Assaf Zeevi at the interview, I immediately knew CBS was the right place for me. Once I was here, I really loved the friendly and supportive environment. The faculty and staff truly care about the students and want them to succeed. Professor Russo and Zeevi, who are my advisors, provided excellent support and guidance to me and were deeply involved in my research. They would actively participate by providing feedback on my data and numerical figures, and even join in on doodling on a whiteboard with me to think out the problem,” says Zhang.
“One of the most exciting aspects of the program is the opportunity to take classes all around Columbia and meeting people of all diverse backgrounds who have different perspectives, yet interested in the similar topics concerning AI and ML” Zhang says. Some of his favorite classes include Foundations of Stochastic Modeling, High-dimensional Statistics, and Computational Learning Theory. “The opportunity to take classes outside of CBS really enhanced my learning of the basics and fundamentals needed in my field, with an extra bonus to making friends from other departments.”
Overall, Zhang describes the academic atmosphere of the DRO community to be incredibly collaborative. “Every time there is a seminar, we have a “before-party” to discuss the speaker’s paper in advance. I also look forward to our brown bag lunch events, where students and faculties would present their research and exchange ideas, and the smoked salmon focaccia with herbed cream cheese is amazingly tasty!”
“I can say with 100% confidence that the DRO program has a prestigious operations research group and it is incredibly easy to maintain strong and close relationships with other departments like IEOR, Statistics, Computer Science, and the Data Science Institute, while also fostering close collaboration with connections from the healthcare, technology, and financial industries.”