The self-similar evolution of stationary point processes via persistent homology

Daniel Spitz,Anna Wienhard
2023-08-04
Abstract:Persistent homology provides a robust methodology to infer topological structures from point cloud data. Here we explore the persistent homology of point clouds embedded into a probabilistic setting, exploiting the theory of point processes. We introduce measures on the space of persistence diagrams and the self-similar scaling of a one-parameter family of these. As the main result we prove a packing relation between the occurring scaling exponents.
Probability
What problem does this paper attempt to address?