Resonant anomaly detection without background sculpting

Abstract

We introduce a new technique named Latent CATHODE (LaCATHODE) for performing "enhanced bump hunts", a type of resonant anomaly search that combines conventional one-dimensional bump hunts with a model-agnostic anomaly score in an auxiliary feature space where potential signals could also be localized. The main advantage of LaCATHODE over existing methods is that it provides an anomaly score that is well behaved when evaluating it beyond the signal region, which is essential to prevent the sculpting of background distributions in the bump hunt. LaCATHODE accomplishes this by constructing the anomaly score directly in the latent space learned by a conditional normalizing flow trained on sideband regions. We demonstrate the superior stability and comparable performance of LaCATHODE for enhanced bump hunting in an illustrative toy example as well as on the LHC Olympics R&D dataset.

11 pages, 8 figures; v2 (published version): referencing code and minor style updates

References in corpus (2)

• PYTHIA 6.4 Physics and Manual
• Extending the Bump Hunt with Machine Learning

Cited by in corpus (24)

• How to Understand Limitations of Generative Networks
• The MadNIS Reloaded
• The Interplay of Machine Learning--based Resonant Anomaly Detection Methods
• Deep Generative Models for Detector Signature Simulation: A Taxonomic Review
• Nanosecond anomaly detection with decision trees and real-time application to exotic Higgs decays
• Precision-Machine Learning for the Matrix Element Method
• A Method to Simultaneously Facilitate All Jet Physics Tasks
• Anomalies, Representations, and Self-Supervision
• Normalizing Flows for High-Dimensional Detector Simulations
• Combining Resonant and Tail-based Anomaly Detection
• Convolutional L2LFlows: Generating Accurate Showers in Highly Granular Calorimeters Using Convolutional Normalizing Flows
• Differentiable MadNIS-Lite
• Anomaly detection with flow-based fast calorimeter simulators
• Non-resonant Anomaly Detection with Background Extrapolation
• Unifying Simulation and Inference with Normalizing Flows
• Foundations of automatic feature extraction at LHC--point clouds and graphs
• Weakly supervised anomaly detection for resonant new physics in the dijet final state using proton-proton collisions at $\sqrt{s}=13$ TeV with the ATLAS detector
• Inferring correlated distributions: boosted top jets
• Data-Driven High-Dimensional Statistical Inference with Generative Models
• Finding excesses in model parameter space
• Observable Optimization for Precision Theory: Machine Learning Energy Correlators
• Weakly supervised anomaly detection with event-level variables
• Strong CWoLa: Binary Classification Without Background Simulation
• Graph theory inspired anomaly detection at the LHC