Parallel multi-scale reduction of persistent homology filtrations

TL;DR

This paper introduces a parallel algorithm for reducing boundary matrices in persistent homology, significantly speeding up computations and enabling simultaneous multi-scale barcode estimation.

Contribution

It extends previous dependency-based reduction methods by proposing a novel parallel algorithm that reduces computational cost and allows multi-scale barcode analysis.

Findings

Reduces computational iterations to about 0.1% of traditional methods.

Enables simultaneous multi-scale barcode estimation.

Achieves high precision in fewer iterations.

Abstract

The persistent homology pipeline includes the reduction of a, so-called, boundary matrix. We extend the work of Bauer et al. (2014) and Chen et al. (2011) where they show how to use dependencies in the boundary matrix to adapt the reduction algorithm presented in Edelsbrunner et al. (2002) in such a way as to reduce its computational cost. Herein we present a number of additional dependencies in the boundary matrices and propose a novel parallel algorithms for the reduction of boundary matrices. In particular, we show: that part of the reduction is immediately apparent, give bounds on the reduction needed for remaining columns, and from these give a framework for which the boundary reduction process can be massively parallelised. Simulations on four synthetic examples show that the computational burden can be conducted in approximately a thousandth the number of iterations needed by traditional methods. Moreover, whereas the traditional boundary reductions reveal barcodes sequentially from a filtration order, this approach gives an alternative method by which barcodes are partly revealed for multiple scales simultaneously and further refined as the algorithm progresses; simulations show that for a Vietoris-Rips filtration with $\sim10^4$ simplices, an estimate of the essential simplices with 95% precision can be computed in two iterations and that the reduction completed to within 1% in about ten iterations of our algorithm as opposed to nearly approximately eight thousand iterations for traditional methods.