Below, you'll find information and tips for applying to join our research group at Mila.
Your tuition fees are covered. The stipend for MSc students is 24K CAD/year and for PhD students, 27K CAD/year.
What topics does your lab work on?
My most recent publications provide a good sense. In sum, we work on (i) causal inference from large-scale data using ML methods, and (ii) applying causality to make ML more generalizable.
Applications are due Dec 1, 2022. Choose my name from among the list of supervisors.
Can I email you directly?
I discourage cold emails that generically name one of my papers, and list lots of your projects. These generic emails rarely help.
Cold emails may only help if: you have research experience with causality-related topics, and you mention that experience directly and succinctly.
What are ways to have a compelling application package?
There are many paths to having a great application, but there are a few notable things I look for when I review applications:
In general: I look for indications in the application that a candidate is mature in their thinking and problem solving, driven by a vision or goal, and intellectually curious. I do my best to read between the lines for these signs. For me, it’s more important that you’re an independent thinker than that you list several of my papers in your application, especially if you haven’t read them deeply.
Statement: I pay attention when candidates describe their research vision – even if the topic isn’t one of my core research topics – with clarity and depth. (Of course, if the topic is too far from my research area, I’ll still be impressed but unable to move forward.) In a short statement, low-level details aren’t feasible but if you can convey the problems you want to study and your motivations for it, going beyond generic keywords, that shows me that you are self-driven. I also especially note candidates that already have background or research experience with causality and probabilistic models – feel free to bold such statements.
Transcripts: I look at grades in math classes like probability and statistics and linear algebra, and CS classes like algorithms and data structures. If you didn’t have great grades initially but improved, explain why in your statement. A thoughtful explanation of your personal growth goes a long way.
What will the 1:1s be like?
Based on applications, I select anywhere from 10 to 15 candidates to (virtually) meet 1:1. These 1:1s will not be a surprise: I email the candidates at least one week prior to the 1:1s and describe how the meeting will be structured. The goal of this is to give you as fair a chance to prepare as possible.
I do vary the structure of these 1:1s from year-to-year so I can’t say exactly what this year’s will be like, but I can say that I do ask some technical questions. In my prep email, I include topics that I might ask about.
Are there any particularly important things you look for in 1:1s?
Besides the things I noted above that I look for in applications, in 1:1s, I put a lot of emphasis on communication. This encompasses everything from clarity (how you describe your background and your research vision), how well we communicate to each other, and what your communication style is. In my lab, we stress openness and inclusiveness and I evaluate a candidate’s fit for this environment.