Speaker: Abhinav Chand, Kansas State University

Title:  Applications of graph neural networks to modulus on networks

Abstract: Inspired by recent applications of machine learning to mathematics, we trained graph neural networks to predict the modulus on graphs. We used saliency analysis to identify important features of the graph that affect the minimum feasible partition problem. We believe this could be applied to other problems in network science/graph theory to identify patterns and generate conjectures.

Location: Burt 204 (or join by Zoom)