Decomposing zero-dimensional persistent homology over rooted tree quivers

TL;DR

This paper studies the decomposition of zero-dimensional persistent homology over rooted tree posets, providing a finite type classification and an efficient algorithm for decomposing representations.

Contribution

It characterizes the additive closure of zero-dimensional persistent homology representations over rooted trees and offers a quadratic-time decomposition algorithm.

Findings

Additive closure is of finite type for rooted tree posets.

Provides a quadratic-time algorithm for decomposition into indecomposables.

Connects the representations to Ringel's tree modules and Kinser's reduced representations.

Abstract

Given a functor from any category into the category of topological spaces, one obtains a linear representation of the category by post-composing the given functor with a homology functor with field coefficients. This construction is fundamental in persistence theory, where it is known as persistent homology, and where the category is typically a poset. Persistence theory is particularly successful when the poset is a finite linearly ordered set, owing to the fact that in this case its category of representations is of finite type. We show that when the poset is a rooted tree poset (a poset with a maximum and whose Hasse diagram is a tree) the additive closure of the category of representations obtainable as zero-dimensional persistent homology is of finite type, and give a quadratic-time algorithm for decomposition into indecomposables. In doing this, we give an algebraic characterization of the additive closure in terms of Ringel's tree modules, and show that its indecomposable objects are the reduced representations of Kinser.