Reeves Receives NSF CAREER Award to Advance Theoretical Underpinnings of Data Science

Galen Reeves, assistant professor of electrical and computer engineering and statistical science at Duke University, has been awarded a prestigious National Science Foundation Faculty Early Career Development (CAREER) Award. The award supports outstanding young faculty members in their efforts to build a successful research enterprise. For the next five years, the $490,000 grant will help Reeves explore the limits of emerging data analytics tools such as machine learning and neural networks used in signal processing, machine learning and statistics. Examples include designing new algorithms for inference, learning and compression, as well as analyzing bi-linear and multi-layer inference problems with applications to deep learning..

Reeves’s research takes tools and ideas from a variety of disciplines, such as information theory and statistical physics, and uses them to determine if a problem is solvable or if an existing algorithm is as good as it could be. This research leverages powerful mathematical tools, such as free probability theory and the replica method from statistical physics, as well as recent algorithmic breakthrough based on approximate message passing algorithms.

Understanding how data processing tools work and what their limits are can have large impacts on fields such as wireless communications, fiber optic communications and computational photography. Reeves’s research can also help reveal how complex algorithms such as deep neural networks work, helping to make them more efficient and applicable to more types of datasets and computational problems.

“The big challenge in my research is learning how these systems actually work, so you can train them better, broaden the scope of problems they can be used on, and transfer things learned from one setting to another,” said Reeves. “There’s a ton of work being done on these systems, and I think theoretical contributions can have a massive impact at this point in time.”