Discovering New Mathematical Structures with Machine Learning

Explore a 59-minute lecture from the CMSA Mathematics and Machine Learning Closing Workshop where Professor Kyu-Hwan Lee from the University of Connecticut demonstrates how machine learning can be leveraged to uncover novel mathematical structures. Delve into two fascinating case studies examining murmurations in number theory and the loadings of partitions connected to Kronecker coefficients in representation theory and combinatorics. Learn about an innovative approach that focuses on analyzing mathematical objects collectively rather than individually, creating robust datasets specifically designed for machine learning experiments and their subsequent interpretation.