Computing the multicover bifiltration

From MaRDI portal
(Redirected from Publication:6174807)
Computing the multicover bifiltration (scientific article; zbMATH DE number 7729238)



Abstract: Given a finite set AsubsetmathbbRd, let Covr,k denote the set of all points within distance r to at least k points of A. Allowing r and k to vary, we obtain a 2-parameter family of spaces that grow larger when r increases or k decreases, called the emph{multicover bifiltration}. Motivated by the problem of computing the homology of this bifiltration, we introduce two closely related combinatorial bifiltrations, one polyhedral and the other simplicial, which are both topologically equivalent to the multicover bifiltration and far smaller than a v Cech-based model considered in prior work of Sheehy. Our polyhedral construction is a bifiltration of the rhomboid tiling of Edelsbrunner and Osang, and can be efficiently computed using a variant of an algorithm given by these authors. Using an implementation for dimension 2 and 3, we provide experimental results. Our simplicial construction is useful for understanding the polyhedral construction and proving its correctness.


This paper describes a method of computing properties of a bifiltration. Bifiltrations are natural extensions of the standard filtration in TDA. They apply this to some data sets, and look at natural generalizations of the 1-D case of the offset filtration (e.g. Čech). They introduce the bifiltrations ``Cov, ``S-Del, ``S-Rhomb, and ``Rhomb and show that these bifiltrations are weakly equivalent. The authors also discuss the size of the bifiltrations, showing that the size bound is not very large given certain conditions (e.g. points being in general position). In particular, this beats the exponential bound in previous work. They then discuss an algorithm that computes ``Rhomb (the rhomboid bifiltration), also providing code to do the computation (currently limited to 2 and 3 dimensions). The code is easily generalizable to higher dimensions, but it becomes computationally expensive. They conclude with some experiments performed on data in \(\mathbb{R}^2\) and \(\mathbb{R}^3\) (e.g. points sampled uniformly at random from the unit interval \([0,1]^2\) or \([0,1]^3\), or a disk, or an annulus). They get promising results, but still need to address how to deal with larger data sets. Instead of truncation, they suggest an inductive process or a coarsened biltration.



Cites work









This page was built for publication: Computing the multicover bifiltration

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6174807)