Swetha Revanur only just graduated from high school in San Jose, California, but has already co-authored a paper in Nature Communications, built a health education app, interned at the National Institutes of Health, and placed first in bioinformatics research at the 2015 and 2016 Intel International Science and Engineering Fairs. Now she’s interning for the summer in HP’s Emerging Compute Lab before entering Stanford University in the fall, where she hopes to major in computer science.
HP: How did you get interested in data and computer science?
When I was a freshman in high school I did a project in my history class about a disease called Hidradenitis suppurativa, which is a skin disease almost no one has heard of but affects an estimated 1% of the global population. I started to look at the genetics behind the disease and came across these huge databases of information and realized that the best way to take advantage of that data was with computer science.
HP: What were you able to make of the data?
I developed algorithms that linked Hidradenitis suppurativa with other related diseases. It involved creating a scoring matrix that can tell us how closely associated any two diseases are. That led me to look at gene co-alterations. We know that diseases like cancer, heart disease, and diabetes, for example, can all occur when multiple genes in our body are simultaneously altered, but we didn’t have a computationally robust method for identifying these co-altered gene pairs. That’s what I was working on for the project I took to the Intel fair this past May.
HP: So what are you working on at HP Labs this summer?
I’m a machine learning intern here, and I’m using statistical models along with some graph theory to generate sensor data for pervasive spaces. There’s this idea of buildings fitted with sensors providing “ambient intelligence” – we want to be able to take advantage of this sensor data to tailor living environments to people’s needs. What I’m doing specifically is using stochastic and probabilistic models to generate synthetic sensor data and creating simulations of how a person would behave in a household. I can then use machine learning to draw conclusions from these simulations about patterns of human behavior that could help us design ambient technologies that optimize energy consumption, aid healthcare, or increase security, for example.
HP: How are you enjoying being at HP Labs?
It’s a great experience. It’s very different from doing basic science research. I feel like I have a lot more freedom to explore here and I like that the work you are doing can have a much more immediate impact. The researchers here are really brilliant, too, so it’s amazing getting to talk to all these different people.
HP: Has being at HP Labs changed how you are thinking about your career?
Sure. A few months ago I was planning on majoring in computer science with a concentration in bio-computation, but since starting here I’ve been thinking that maybe instead I’ll focus on artificial intelligence and machine learning.
HP: What do you like to do for fun?
I actually love to fence – I do sabre fencing. I’m also trained in Indian classical vocals and dance. And then I like to make applications and tools that help address public health problems. Last year I built one related to youth smoking and I’m working on another one right now that’s related to global health and sanitation.