Betti numbers in multidimensional persistent homology are stable functions (Q2857523)
From MaRDI portal
| This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use this page instead for the normal view: Betti numbers in multidimensional persistent homology are stable functions |
scientific article; zbMATH DE number 6222224
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | Betti numbers in multidimensional persistent homology are stable functions |
scientific article; zbMATH DE number 6222224 |
Statements
Betti numbers in multidimensional persistent homology are stable functions (English)
0 references
4 November 2013
0 references
multidimensional persistence
0 references
persistence diagram
0 references
Čech homology
0 references
foliation
0 references
continuous filtering function
0 references
natural pseudo-distance
0 references
persistent Betti numbers
0 references
Define the relation \(\prec\) on \({\mathbb R}^{n}\) by \(u \prec v\) iff \(u_{i} < v_{i}\) for all \(i = 1,\ldots,n\). Given a function \(\phi : X \rightarrow {\mathbb R}^{n}\) for any \(u \prec v\) one has an induced mapping of Čech homology \(\pi_{k}^{(u,v) }: H_{k}(X(\phi \prec u) \rightarrow H_{k}(X(\phi \prec v))\). The image of this map is called the multidimensional \(k\)th persistent homology group of \((X,\phi)\) at \((u,v)\). The authors show that if \(X\) is a triangulable space then these Betti numbers are finite. One can associate to the maps \(\phi\) a diagram in \(\Delta^{+} = \{(u,v) \in {\mathbb R}^{n} \times {\mathbb R}^{n}: u \prec v\}\) that represents the persistent Betti numbers. If the map \(\phi\) is perturbed then the diagram shifts and a distance between the diagrams can be defined. The main result of the paper is that these multidimensional persistent Betti numbers are stable in the sense that small perturbations in the function \(\phi\) induce only small changes in the Betti numbers. This is significant for the applications of persistent homology to the study of large data sets.
0 references