Topology-Informed Jet Tagging using Persistent Homology

TL;DR

This paper introduces a novel jet classification method using persistent homology to extract topological features from particle jet data, enhancing quark-gluon discrimination.

Contribution

It applies persistent homology to particle jets, capturing topological features like loops, and integrates these features into a neural network for improved jet tagging.

Findings

Persistence images from H0 and H1 are comparably effective.

Loop-like topological features encode meaningful jet substructure information.

Topology-informed features improve jet classification performance.

Abstract

We present a topology-informed approach for classifying particle jets using persistent homology, a framework that captures the structural properties of point clouds. Particle jets produced in proton-proton collisions consist of cascades of particles originating from a common hard interaction. Each jet constituent is represented as a point in a three-dimensional feature space defined by the relative transverse momentum and angular coordinates with respect to the jet axis, yielding a point cloud description of each jet. Persistent homology is computed using a Vietoris-Rips filtration to obtain persistence diagrams, which are subsequently converted into persistence images for the H0 and H1 homology groups, corresponding to connected components and loop-like structures, respectively. These persistence images are used as inputs to a convolutional neural network for quark-gluon jet classification. Using the publicly available HLS4ML jet dataset, we find that persistence images derived from H0 and H1 individually exhibit comparable discriminative power. These results demonstrate that loop-like topological features, which are not conventionally exploited in jet tagging, encode meaningful information about jet substructure and provide a promising direction for topology-informed jet classification.