Filtration evolution of hypergraphs: A novel approach to studying multidimensional datasets

The rapid growth of large datasets has led to a demand for novel approaches to extract valuable insights from intricate information. Graph theory provides a natural framework to model these relationships, but standard graphs may not capture the complex interdependence between components. Hypergraphs are a powerful extension of graphs that can represent higher-order relationships in the data. In this paper, we propose a novel approach to studying the structure of a dataset using hypergraph theory and a filtration method. Our method involves building a set of hypergraphs based on a variable distance parameter, enabling us to infer qualitative and quantitative information about the data structure. We apply our method to various sets of points, dynamical systems, signal models, and real electrophysiological data. Our results show that the proposed method can effectively differentiate between varying datasets, demonstrating its potential utility in a range of scientific applications.