Benjamin Choi built a mind-controlled bionic arm as a high school student in Virginia, which led him to realize something important about the human brain. That means your brain is noisy. Your brain is constantly sending out all kinds of signals that regulate everything from your breathing to your hunger to planning your day’s schedule. For his prosthetic leg to work, he needed software that could filter out noise and recognize signals specifically intended for manipulating the arm.
“One of the lessons I learned from that project that really stuck with me was that there can be very noisy, high-dimensional signals that you don’t know what’s going on, but there are all kinds of fun things you can do from linear algebra to actually find real signals in this kind of noisy space,” he said.
This project formed the basis for Choi’s interest in applied mathematics and machine learning. Now in his fourth year, Choi is graduating with a degree in applied mathematics and a master’s degree in computer science. Harvard University John A. Paulson School of Engineering and Applied Sciences (SEAS). He has worked with Johns Hopkins University, NASA, and Kempner Institute for Natural and Artificial Intelligence, Harvard UniversityAfter graduating, you will work in industry as a machine learning researcher.
“What I liked about the applied mathematics curriculum was that it felt similar to the liberal arts in STEM,” Choi said. “I like the interdisciplinary nature of it. I feel like I have a wide range of interests in general, but the flexibility of the applied mathematics curriculum has allowed me to take advantage of that while also delving into some very interesting topics.”
Choi loved mathematics, but when applying for university entrance exams, she did not consider applied mathematics as a possible major. Although Harvard’s broad liberal arts education was appealing to him, it was his first visit to the science and engineering complex that really solidified his decision to come here.
“I remember visiting the Allston campus and being really, really impressed,” he said. “It was clear to me that this was an area where the school was making significant investments, and it definitely paid off in terms of the great classes, faculty, and resources that were available to me while I was here.”
One particularly impactful class was “AM220: Geometric Methods for Machine Learning,” taught by Melanie Weber, assistant professor of applied mathematics and computer science. Choi then joined Weber’s geometric machine learning group as a researcher and eventually became an advisor for his senior thesis comparing how large-scale language models in artificial intelligence map data using a framework similar to human emotions.
“It was a really eye-opening class. It showed me that there are all kinds of really cool intersections between geometry and the fundamental principles that underpin these kinds of machine learning concepts,” he said. “I really, really loved that class. There was a kind of cross-pollination between some of the geometric concepts that I picked up in that class and how I was thinking about my research at the time in signal processing and machine learning. Professor Weber was very kind and very generous in terms of guiding me through some of the research ideas that I had.”
The Kempner Institute began its research efforts just as Choi arrived on campus, giving Choi a second outlet for his machine learning interests. As a junior, he was part of the first cohort of Research experience at Kempner College The students are collaborating with Demba Ba, the Gordon McKay Professor of Electrical Engineering, on a project that applies the same AI architecture behind chatbots to filter out noisy gaps in brainwave signals.
“I was really drawn to the overall theme of this lab, which is whether we can find parallels between artificial intelligence, machine learning, natural intelligence and brain waves, a problem I was already working on with the bionic arm,” he said. “Professor Ba sponsored my research for the first student in the Kempner program. It was a great experience.”
A master’s degree in computer science helped Choi broaden his understanding of machine learning, from the applied mathematics that explains it to the computer architecture that runs it. Throughout her master’s degree, Choi became the lead author of multiple publications and learned both computer science theory and approaches to problem solving that work even when specific programming languages and AI models change over time.
“Harvard’s CS curriculum taught me how to think,” he said. “Teaching them how to think has been really, really valuable and much more useful than learning a particular language in first grade that is no longer used.”
The signal processing that underlies the bionic arm, which he continued to develop at Johns Hopkins University, became the basis of Choi’s interest. But his time at SEAS, whether at the Weber Group or the Kempner Institute, broadened and refined his understanding of machine learning. Mr. Choi is currently looking for a job and is ready to continue what he has been doing. That means finding ways to apply your research to the world around you.
“Throughout my time here, I have sought to gain a variety of experiences in terms of conducting research and applying that research to bring about meaningful change in my field,” he said. “Given Harvard’s place in the Kempner Institute, I really think I’ve had great classes on AI here and heard from all kinds of really great speakers. It’s been really thrilling to see how quickly things are evolving.”
