Justin Solomon, an associate professor in the MIT Department of Electrical Engineering and Computer Science (EECS) and a member of the Computer Science and Artificial Intelligence Laboratory (CSAIL), is employing modern techniques based on ancient geometric principles to solve a wide array of complex problems, specifically in the field of data science.
Drawing upon the geometric fundamentals laid out in antiquity by the likes of Euclid, a Greek mathematician hailed as the ‘father of geometry’, Solomon uses geometry to gain insights from data, asserting that the language commonly used to discuss data – namely, distances, similarities, curvature, and shape – are intrinsically geometric in nature.
The vast range of problems that can be solved geometrically inspired Solomon to establish the loosely-named Geometric Data Processing Group, where approximately half the team focuses on dealing with two- and three-dimensional geometric data. By applying geometric techniques to statistics for advanced AI models, the team is contributing significantly to the field of high-dimensional statistical research.
Solomon’s interest in computer graphics prompted him to double-major in math and computer science at Stanford and subsequently motivate him to intern at Pixar Animation Studios where he developed physical simulations for animations. He continued to work at Pixar throughout college and in graduate school, specializing in realistic animations and rendering techniques to alter the appearance of animations.
After obtaining a PhD in computer science, Solomon decided to pursue an academic career, with his first project focusing on optimal transport, a seemingly niche topic that has a broad range of applications.
Solomon is also committed to making geometrical research accessible to a diverse demographic of students. He launched the Summer Geometry Initiative, a six-week paid research program primarily aimed at undergraduates from underrepresented backgrounds, which has seen significant results. Not only has the program led to a notable change in the composition of incoming PhD students at MIT, but also at other institutions.
His plans for the future include the employment of geometric tools to improve unsupervised machine learning models, often presented with the challenge of recognizing patterns from unlabelled data. Given that labelling 3D data is usually a costly task, Solomon believes models incorporating geometric analysis and data inference can facilitate more effective learning.
Beyond his academic pursuits, Solomon is an enthusiastic musician and currently plays in the New Philharmonia Orchestra in Newton, Massachusetts. He asserts that the analytical nature of music complements his research in computer graphics and aids his creative productivity.
