Even though my interests did not align with typical research internships, I still wanted to go for one as a part of the “IIT experience”. To begin with, I was sure that I didn’t want to apply for one in my core. I was more inclined towards machine learning, so I checked out a few online courses in that domain and moulded my cover letter in that direction. I started mailing a little early and hence I got placed by mid-December. The professor I applied under was in context-based behaviour modelling, and so I mentioned my interest in her field and how I enjoy understanding human nature. She then allocated me to one of her PhD students, and that is how I got an intern at RMIT.
As the pandemic started to take its course, I realized that the situation is not going to clear up soon and hence decided to take up another internship before the RMIT one, hopefully in a business school, since it was more in line with what I wanted to do. By that time I had studied and added more to my knowledge of data analytics, so I started applying for analytics-driven research in IIMs. The professor under whom I did my research did not take me right away, he gave me a problem to work on, which he thought was something we could later work on as a research project. He gave me a week to come up with ideas and combine them into a presentation for him. I contacted a few people and put together a presentation, which he ended up liking and subsequently, offered me the opportunity to work with him
I started my IIM internship around mid-April, and the professor asked me to go through COVID-19 research and collect relevant databases. So for the first 20 days of my IIM intern, I ended up doing just that and exploring machine learning and data analytics. At the end of the first month, we had a review and the professor liked my work. While the first 20 days were quite chill, after the work review things started to pick the pace and the intensity of work increased. We started working on data philanthropy and I spent 5-7 days collecting data and then proceeded to analyze that data to prove hypotheses. By the end of the second month, I was able to identify five indicators to predict data philanthropy nature for users. I presented him with this deliverable, and he felt that it was a good output and so we started working on a research paper together based on that. He gave me additional work which involved predicting the firm's risk exposure to climate change using transcript data from quarterly earning calls. I worked on this with him, even after the end of my internship period. This ended on a high note and we were able to forecast a firm's climate risk exposure, using transcripts of 10,000+ firms with high accuracy.
At RMIT I was interning in meta-learning, a novel machine learning technique to improve upon already existing Ubiquitous computing methods used in the field of Smart Healthcare. I had exposure to machine learning, but meta-learning was something that I learned during this internship. My professor gave me a few research papers to go through. I restructured datasets and extracted relevant features to be used for predictions followed by exploratory data analysis. We were able to predict human stress levels to great accuracy using wrist and chest based device readings.
Overall it was a really good learning experience. After my internship, I started appreciating research work a lot more. So, this helped remove my apprehensions concerning the research domain
Try to narrow down to a domain that you are interested in and start mailing early. I feel that a lot of people are skill-building driven and hence, they keep trying to upskill and put mailing on the back burner. My advice is that once you have identified your interests and have started working on the skills required, start mailing to professors in that field for your research. After securing the intern you will have ample time to build the relevant skills.
Interviewed by Aashita Gupta.