Meet Our Team: Abhisek Gupta
What makes 84.51˚ such a great place to work? Is it the free food regularly found around the office, the challenging problems we work on, or the skylight on our top floor? While those things are certainly great, it’s the people that make working here truly special. To share more about who we are, the *Meet Our Team* series will highlight some unique traits of our team members through one-on-one interviews.
To start off this series, data scientist Abhisek Gupta shares the story of how he got into the field in an interview with fellow data scientist Corinne Schlachter.
Corinne Schlachter (CS): Where did you grow up?
Abhisek Gupta (AG): I grew up in Shahjahanpur which is a small town in the Northern Indian state of Uttar Pradesh.
CS: What is life like in North India?
AG: Life was kind of slow in the town and the whole community in my neighborhood was like a big family. I remember not having a computer until I was 12, which kind of helped because I spent a lot of time my time playing cricket and soccer, flying kites, and playing all kind of board games. Despite all that, there was a still lot of focus on doing well in academics, and it helped greatly when I went to college.
CS: What did you study in undergrad?
AG: I studied civil engineering back in undergrad but became interested in analytics in my junior year.
CS: What made you decide to become a data scientist?
AG: I came across a Coursera course on R programming through Johns Hopkins University and decided to complete it. Switching from the construction field to coding in front of a computer most of the day has been a big change. But I do enjoy every minute of it; trying to model human behavior through mathematical frameworks is pretty inspiring and interesting.
CS: What was your favorite class in grad school?
AG: Statistics 101, Optimization Modeling, and Predictive Analytics were my favorite classes. Statistics helped me understand the impact and usage of stats in daily our lives, and I saw the field come alive in real world examples and that deepened my understanding of first principles. Predictive analytics built on Stats 101 to go more deeply into different forms of supervised and unsupervised learning. It helped me understand that if enough data is there, mathematical models work very well in predicting the future. Optimization modeling took these courses and put them in the context of real world business problems. For example, knowing future house prices is good, but being able to manipulate features of your own house to maximize future value is even better. Optimization modeling actually helped me solve these problems.
CS: Tell us about a project at 84.51˚ in which you applied something learned in grad school.
AG: I employed a lot of my newly learnt skills from Purdue in identifying potential ClickList customers via a lookalike model. The content I studied in Predictive Analytics on supervised classification techniques was central to solving this problem. Additionally, peripheral but important concepts I learned in other courses – like data preparation, identifying anomalous observations, feature selection, etc. – were also very relevant.
CS: What is a data science/statistics topic that you want to learn more about?
AG: I am intrigued by knowledge discovery through pattern mining in literature. It’s no secret that all sciences have exploded in terms of niche research in the last few decades. Now, couple that with the realization that some of the biggest challenges of our time can be solved only through interdisciplinary research. For example, it’s not humanly possible to read thousands of papers in multiple fields to search for associations, but it is indeed possible for an algorithm to do this and identify patterns for us. This kind of research can help us find solutions to problems like disease control, climate change, and more.