## What I Care About Most in Research?

I am Xiaoqi Tan, an Assistant Professor at the University of Alberta and also a Fellow of the Alberta Machine Intelligence Instittue (Amii). I study optimization and decision-making under uncertainty, with a particular focus on the theory of online algorithms and its implications for problems at the interface of multi-agent systems, economics and computation (e.g., auctions; mechanism design; game theory; resource allocation; online markets; economics of clouds/networks).

I enjoy the beauty of mathematics, and consider a technically-sound and aesthetically-elegant theorem the core of publication-worthy results. I admire good applied research, but also believe that not all research needs to have practical use — good theory sustains over time, and may unexpectedly unleash its power. I am most excited about connecting theory with practice to obtain sharp insights into real-world systems design, implementation, and operation. To do so, I believe it is important, and also necessary, to sacrifice certain details of the problem. “I want to know how God created this world. I am not interested in this or that phenomenon. I want to know His thoughts, the rest are details.” — Albert Einstein

## What I Care About Most in Prospective Students?

I often receive emails from prospective students with descriptions like “I know how to apply $X$ to solve $Y$” — this, unfortunately, is not what I care about the most. I care about whether you can convert a real-world problem into rigorous mathematical models, and then develop algorithms to solve the problem with provable guarantees — in the form of mathematical theorems and lemmas. In short, I look for prospective students who are motivated and excited about creating new and rigorous knowledge — to argue “how and why things work.”

## What is My Take on Good Advisor-Advisee Match?

I respect scholarship and love research, like most academics do. I consider getting a graduate degree takes initiative and commitment — it requires strong motivation to excel, long-lasting enthusiasm in research, and probably most importantly, a good advisor-advisee match — based on mutual trust and respect, effortless communication, and sometimes, a bit of luck. While it is complex to define what is exactly a “good match,” a simple rule of thumb is: if you feel this is the person you are willing to “work with,” not to “work for,” then it is usually a good sign.