Motivation
I thought that the process of applying for a research internship was a very hit-and-miss task, and hence decided that if I was going to apply for a research project, I'd do it via a research program. When I didn't make it to those of foreign universities, I realised that SURA was a great option as well, simply because it was a research program like others, and because the quality of research in my department was (in my opinion) comparable to most foreign universities anyway. Additionally, having completed a statistics course in my fourth semester, I knew I wanted a field that was closely related to Statistics and Machine Learning.
Process and Application
I then mailed 5-6 professors; some of them wanted extremely high time commitments, and some had prior experience as a prerequisite. Then, I mailed Prof. Rahul Garg, who told me to give his research paper a read, and things started from there. I speedran through a higher level course he was taking, wrote up a proposal for the project and mailed it to the professor. After his approval, I got it signed by multiple people, and submitted it to the committee. Then, the SURA shortlist was released, and me and my partner began to work on the project.
Experience
The project itself entailed being a part of a research team and working with them on a project, for example, a paper, a project, or an industry collaboration. The team consisted of PhDs, postdocs, and students from other universities as well. Personally, I worked on two projects, both under the SURA umbrella, and was an active member of the team, working mostly on the computational side of things. Our professor had developed a new technique called Temporal Synchronization Analysis. Both the projects I worked on entailed using this technique and statistics to reason differently about brain structures as compared to how conventional models currently do. During the course of the project, we had weekly meetings, where we discussed our current progress, and future line of actions. We used to drop into our professor’s office whenever we had any doubts or new findings to share. There was no fixed time commitment as such, there were extremely active periods and some lukewarm periods throughout the course of the project.
As for the research involved, we dealt with something called FMRI scans, which pass your brain through a high-intensity magnetic field. MRI is to an image, what FMRI is to a movie. It’s akin to a 3D movie monitoring the brain’s activity. You do this after a fixed time, say taking 60 scans, each after one second, for the course of a minute. Then to prove any given hypothesis, you plot the data corresponding to the control and experimental groups and compare them. Then you use mathematical and statistical techniques to reason about this data. The volume of the data we had to process made it challenging too. The project is still ongoing, it’s been 6 months since I’ve been working on the project. I wanted some amount of research experience, and this project has helped me funnel towards certain areas of research in CS.
Advice
For those looking for a SURA project, I'd advise you to have a solid gradesheet, and sit through some of the 700/800 level courses that the professor you wish to work under are taking, so that you can interact with them in a classroom setting and gauge if the material sufficiently interests you. However, don't take SURA simply because your peers are doing it, or because you need something to do over the Summer; do it if you're genuinely interested in research. A SURA is reflective of how actively you took part in research, consequently, more active members would be given higher quality work. And if you do decide to pursue SURA, stick with it. Research can be particularly hard at times, but it’s important to persevere through
Interviewed by - Kavya Chopra