This $193,155 three-year Project Grant from the National Science Foundation's Mathematical and Physical Sciences program (CFDA 47.049) supports research at Brigham Young University to develop a mathematical framework describing how sparse network structures can effectively process information and aggregate it in ubiquitous real-world network patterns. The project aims to advance understanding of how network topology impacts machine learning algorithms' ability to learn from data, starting with an analysis of Reservoir Computers. Outcomes will include new methods for building more accurate and cost-effective Reservoirs using extremely sparse networks, with the goal of generalizing principles to broader machine learning. The award also supports education of graduate and undergraduate students in applied mathematics at the intersection of dynamics, machine learning and network science through mentored research experiences.
Generated 1/6/24, 9:17 PM