Unsupervised clustering for collider physics

TL;DR

This paper introduces UCluster, an unsupervised clustering method using neural network embeddings to categorize collision events in particle physics, aiding in multiclass classification and anomaly detection for new physics searches.

Contribution

The paper presents a novel neural network-based clustering approach tailored for particle physics data, enabling unsupervised event categorization and anomaly detection.

Findings

Effective clustering of collision events demonstrated

Potential for discovering new physics phenomena

Versatile application to classification and anomaly detection

Abstract

We propose a new method for Unsupervised clustering in particle physics named UCluster, where information in the embedding space created by a neural network is used to categorise collision events into different clusters that share similar properties. We show how this method can be applied to an unsupervised multiclass classification as well as for anomaly detection, which can be used for new physics searches.