Cell Complex Neural Networks

Explore the concept of Cell Complex Neural Networks in this 46-minute lecture presented by Mustafa Hajij for the Applied Algebraic Topology Network. Delve into the world of cell complexes, topological spaces constructed from simple blocks called cells, and discover how they generalize graphs, simplicial complexes, and polyhedral complexes. Learn about a novel, combinatorial, and unifying construction for performing neural network-type computations on cell complexes. Examine the introduction of inter-cellular message passing schemes that consider the topology of the underlying space. Understand how this method generalizes many popular types of graph neural networks, providing a comprehensive framework for computational analysis on complex topological structures.