My life has become far less hectic now that the ICML deadline has passed. Academic research is an interesting beast. From the outside looking in, it often seems like little to no work is being done until just a few weeks before a deadline. That frantic rush to get a publishable result during those critical weeks is a time-honored hallmark of academic life.
But before everyone condemns academics as being extreme procrastinators, consider that the problems being tackled are very open ended and often very difficult to reason about. Indeed, part of what makes academia different from industry is a focus on developing frameworks with which to reason about various phenomena. It's very difficult to constantly bombard one's brain with such problems.
Due to this focus on general applicability, it's often in a researcher's interest to focus on ideas which impact many different areas. As such, it's important to first understand the key components of the problem at hand before blindly diving into a coding frenzy. I spent a good portion of the semester trying out different approaches, running into dead ends and generally letting the problem marinate in my brain. It wasn't until about six weeks before the conference deadline that I found something I was reasonably sure would work. After that, of course, comes the mad rush to tie up loose ends in the algorithm and code it up so I can run experiments.
In other news, I'm happy to announce that I've been awarded a Microsoft Fellowship. The process was pretty intense. Each department can only nominate three students to apply for the fellowship, and my department took a while on deciding who to nominate. Consequently, I only discovered a few days before the deadline that I was nominated. To make matters worse, my school was on an extended break at that time. What ensued was a mad scramble to contact my references that ended with my last reference replying to me mere hours before the deadline. I also spent a good deal of what was supposed to be a relaxing break writing a research proposal. Many thanks go to Parisa, who offered plenty of constructive criticism.
The fellowship selection process has two stages. 56 finalists (from about 140 applicants) were flown to Redmond to interview with Microsoft researchers. I had seven interviews spanning almost eight hours. It felt almost like a job interview, although I hear those take multiple days. The experience was a bit mind-numbing since the interview day happened to fall into that critical time leading up to the ICML deadline.
I'll be flying to Redmond in a few weeks to attend the awards ceremony. In addition to covering my funding, I will also receive a tablet PC with the latest and greatest Vista technology. Seeing as how I don't really need so many machines running Windows, I'll probably convert my desktop into a Linux box (about damn time, some of my friends would probably say).
Life will get soon get busy again. The Cornell Glee Club has quite a few events lined up in the next few weeks, including a China Tour over Spring break. I've also made it onto my first program committee for ECML 2008, which is probably more exciting to me than it should be. Well, that's all for now, folks. Rehearsal is starting soon.