To Prospective Students

If you’ve found this page, you are likely interested in our research—welcome! I am always happy to hear from students at all levels (PhD, Master’s, and undergrads) who are curious about opportunities of working together.

Our group focuses on developing principled and rigorous approaches to machine learning, strategic modeling and related areas. While I do not believe there is a single “profile” for a perfect student, the following qualities tend to make our collaboration especially productive:

(1) Curiosity and self-motivation.
I believe a good research should and can only be driven by genuine interest. I encourage students to develop their own questions and directions rather than simply following assigned tasks. Typically, we will work together to iteratively refine ideas through reading and discussions.

As a faculty member, I engage in multiple projects at one time. Therefore, the research progress really hinges on your own initiative and engagement – that said, you will not be productive and able to learn much if you lack a strong motivation or cannot devote much time.

(2) Mathematical maturity.
Much of our work is theory-oriented. A solid foundation in linear algebra, calculus, probability and statistics is important for engaging deeply with the project. And of course, additional background in more advanced mathematics (e.g., stochastic process, game theory, real analysis, dynamical systems, etc) is always welcome.

(3) Programming experience.
Efficiency with Python and the experience of implementing and experimenting with machine learning pipelines is helpful. While vibe coding can largely accelerate the process, it remains important to understand and verify what you build.

Depending on your background, you may find the following guidance helpful:

For UNC STOR PhD Students

If you are considering me as a potential thesis advisor, I strongly recommend reaching out during your first year. I usually work with students for at least one semester before making a long-term advising commitment.

I understand that the first year of our PhD program is primarily about coursework and passing CWE, so I do not expect immediate research results. Reaching out early is NOT meant to be a test. Rather, I see it as an opportunity for regular interaction—through informal discussions, paper reading, and exploratory ideas—which helps us better understand whether our research interests and working styles align. This mutual understanding is important for making thoughtful and responsible decisions on both of us.

For UNC Undergraduate and Master’s Students

Due to limited capacity, I typically only work with students who have taken my courses and demonstrated impressive works and engagement there.

However, in case you do not have the chance to take my course, you are still welcome to reach out with your CV and transcripts. A helpful way to start is to read one or two of my papers and share what you found interesting, along with questions or directions you would like to explore.

For Prospective PhD Applicants

Please apply through the UNC STOR PhD program and indicate your interest in working with me. You are welcome to email me with your CV and transcripts, and briefly explaining why your background and interests align with my research. I’ll try my best to respond to relevant ones.

For Visitors and Interns

I will not be hosting in-person summer interns in 2026 due to travel. However, remote collaboration may be possible if there is a strong research fit.

More broadly, I see advising as a collaborative process rather than a top-down management process. I respect each student’s intellectual independence, and my goal is to support you in developing your own research identity while providing guidance and feedback along the way. If this approach resonates with you, I would be glad to connect.