Prospective PhD Students

I am always looking for strong PhD students. Students interested in working with me can apply to BU's PhD program in Computing & Data Sciences (CDS). All applications must be received this way. Questions may be directed to cds-admissions@bu.edu. Please note that PhD admissions are done centrally by committee at the department level, and not by specific faculty members.

I am also affiliated with Computer Science and thus can co-advise students if they apply through BU's PhD program in Computer Science (CS). However, there are many advantages to CDS, so I only recommend applying through CS to work with me if you are also interested in working with the other CS faculty as well and are interested in being co-advised.

Funding: BU CDS and CS students do not pay tuition and when funded, get paid a stipend for their work in the PhD programs. All CDS and CS PhD students are guaranteed funding via either research assistantships (RAs) or Teaching Assistantships (TAs) for 5 years. CDS students receive departmental funding for their first two years to have the freedom to take courses and rotate with advisors if they would like. For more information about CDS, see my blog post here.

Academic background: I enjoy working with people from diverse academic backgrounds. An undergraduate degree in math or computer science is useful but is not necessary. Rather than specific coursework or degrees, I believe the most important attributes in prospective students are mathematical maturity, initiative, and the ability to self-teach oneself material as needed—with these traits, a student of any background can succeed in working with me. Resources for learning relevant background can be found on my resources page.

Resources for applying:

Mor Harchol-Balter's guide to applying to CS PhD programs (last updated 2014)
A number of programs were initiated last year giving feedback on applications, e.g., Stanford Student-Applicant Support Program, Columbia Pre-submission Application Review Program, UW Pre-Application Mentoring Service—take a look around for these.

My research interests: I am generally interested in algorithmic approaches and how they intersect with strategic behavior and incentives, as in the paradigm of mechanism design. My two major research thrusts are in mechanism design for social good (see a high-level talk or a more technical talk) and in rethinking standard behavioral and informational assumptions as they pertain to mechanism design (e.g., robustness to informational interdependence, complementarities, and risk-aversion). I also have extensive prior work in multi-item revenue maximization, as in the <+interdimensional setting.

Research projects: I do not hire graduate students to work on existing research projects. I find research projects that are of mutual interest to myself and the graduate student to work on, such as in the areas of described above. I am also very excited to work with students who have their own questions if the area intersects with my interests.

Undergraduate Research

Due to overwhelming demand and limited availability, I only work with BU undergrads. If you think your case is an exception, please have it arranged through your current research advisor.

What does undergraduate research with me look like?

Research with me looks like a one-on-one theoretically-based project. Sometimes, for undergraduates, you may take on more coding or simulations to explore questions related to a theoretical project, but you are still expected to understand the theoretical background and the overarching theoretical questions that we are aiming to address and drive the project. Otherwise, projects are purely theoretical, as if doing homework for DS 320 but without the sub-questions laid out or the answers already known. Because of (a) the theoretical proficiency required for these projects, (b) how the project is centered around the undergraduate student, and (c) how involved I am in these projects, I take on at most 2 undergraduate students per year, and I expect them to have already taken DS 320 and DS 574 (expect in rare exceptions for the latter) and performed very well in the courses.

Note that this is not necessarily what research looks like in different areas or with different faculty. In some labs, undergraduates are more a cog in a greater machine, requiring less prior expertise or faculty involvement. I encourage students generally looking for research experience (but not specifically in theoretical computer science or algorithmic game theory) to also look for other opportunities, such as with other faculty and BU Spark!.