Separated at Birth: Jet Maximization, Axis Minimization, and Stable Cone Finding

TL;DR

This paper unifies three different jet-finding methods in high-energy physics as variants of a single optimization problem, showing their equivalence for cone jets and analyzing their differences in practical collider scenarios.

Contribution

It demonstrates that three distinct jet-finding approaches are fundamentally connected through a common optimization framework, especially for fixed cone jets.

Findings

The three jet-finding methods are identical for fixed cone jets when properly defined.

Differences between methods are mild for small radius jets in collider experiments.

A unified perspective simplifies understanding and comparing jet algorithms.

Abstract

Jet finding is a type of optimization problem, where hadrons from a high-energy collision event are grouped into jets based on a clustering criterion. As three interesting examples, one can form a jet cluster that (1) optimizes the overall jet four-vector, (2) optimizes the jet axis, or (3) aligns the jet axis with the jet four-vector. In this paper, we show that these three approaches to jet finding, despite being philosophically quite different, can be regarded as descendants of a mother optimization problem. For the special case of finding a single cone jet of fixed opening angle, the three approaches are genuinely identical when defined appropriately, and the result is a stable cone jet with the largest value of a quantity J. This relationship is only approximate for cone jets in the rapidity-azimuth plane, as used at the Large Hadron Collider, though the differences are mild for small radius jets.