Challenges for Unsupervised Anomaly Detection in Particle Physics

Abstract

Anomaly detection relies on designing a score to determine whether a particular event is uncharacteristic of a given background distribution. One way to define a score is to use autoencoders, which rely on the ability to reconstruct certain types of data (background) but not others (signals). In this paper, we study some challenges associated with variational autoencoders, such as the dependence on hyperparameters and the metric used, in the context of anomalous signal (top and $W$) jets in a QCD background. We find that the hyperparameter choices strongly affect the network performance and that the optimal parameters for one signal are non-optimal for another. In exploring the networks, we uncover a connection between the latent space of a variational autoencoder trained using mean-squared-error and the optimal transport distances within the dataset. We then show that optimal transport distances to representative events in the background dataset can be used directly for anomaly detection, with performance comparable to the autoencoders. Whether using autoencoders or optimal transport distances for anomaly detection, we find that the choices that best represent the background are not necessarily best for signal identification. These challenges with unsupervised anomaly detection bolster the case for additional exploration of semi-supervised or alternative approaches.

arXiv comment22 + 2 pages, 8 figures, 2 tables

References in corpus (70)

• Adam: A Method for Stochastic Optimization
• The automated computation of tree-level and next-to-leading order differential cross sections, and their matching to parton shower simulations
• The anti-k_t jet clustering algorithm
• An Introduction to PYTHIA 8.2
• FastJet user manual
• Variational Inference: A Review for Statisticians
• DELPHES 3, A modular framework for fast simulation of a generic collider experiment
• An Introduction to Variational Autoencoders
• Dispelling the N^3 myth for the Kt jet-finder
• Variational Inference with Normalizing Flows
• Identifying Boosted Objects with N-subjettiness
• Anomaly Detection with Density Estimation
• Searching for New Physics with Deep Autoencoders
• Anomaly Detection for Resonant New Physics with Machine Learning
• Learning New Physics from a Machine
• Extending the Bump Hunt with Machine Learning
• Jet-Images: Computer Vision Inspired Techniques for Jet Tagging
• Improving Variational Inference with Inverse Autoregressive Flow
• How to GAN LHC Events
• A generic anti-QCD jet tagger
• QCD or What?
• Simulation Assisted Likelihood-free Anomaly Detection
• Novelty Detection Meets Collider Physics
• Adversarially-trained autoencoders for robust unsupervised new physics searches
• Dijet resonance search with weak supervision using $\sqrt{s}=13$ TeV $pp$ collisions in the ATLAS detector
• Pulling Out All the Tops with Computer Vision and Deep Learning
• Guiding New Physics Searches with Unsupervised Learning
• Variational Autoencoders for New Physics Mining at the Large Hadron Collider
• The Metric Space of Collider Events
• The Dark Machines Anomaly Score Challenge: Benchmark Data and Model Independent Event Classification for the Large Hadron Collider
• Sylvester Normalizing Flows for Variational Inference
• Uncovering latent jet substructure
• Learning to Classify from Impure Samples with High-Dimensional Data
• Tag N' Train: A Technique to Train Improved Classifiers on Unlabeled Data
• Autoencoders for unsupervised anomaly detection in high energy physics
• Anomaly detection with Convolutional Graph Neural Networks
• Weakly Supervised Classification in High Energy Physics
• Finding New Physics without learning about it: Anomaly Detection as a tool for Searches at Colliders
• Better Latent Spaces for Better Autoencoders
• Variational Autoencoders for Anomalous Jet Tagging
• Simulation-Assisted Decorrelation for Resonant Anomaly Detection
• A robust anomaly finder based on autoencoders
• Bump Hunting in Latent Space
• Exploring the Space of Jets with CMS Open Data
• (Machine) Learning to Do More with Less
• Rare and Different: Anomaly Scores from a combination of likelihood and out-of-distribution models to detect new physics at the LHC
• Lund jet images from generative and cycle-consistent adversarial networks
• Learning the latent structure of collider events
• Quasi Anomalous Knowledge: Searching for new physics with embedded knowledge
• Unsupervised clustering for collider physics
• The Hidden Geometry of Particle Collisions
• Comparing Weak- and Unsupervised Methods for Resonant Anomaly Detection
• Deep Set Auto Encoders for Anomaly Detection in Particle Physics
• Mass Unspecific Supervised Tagging (MUST) for boosted jets
• Use of a Generalized Energy Mover's Distance in the Search for Rare Phenomena at Colliders
• Transferability of Deep Learning Models in Searches for New Physics at Colliders
• Unsupervised Outlier Detection in Heavy-Ion Collisions
• The LHC Olympics 2020: A Community Challenge for Anomaly Detection in High Energy Physics
• Linearized Optimal Transport for Collider Events
• Anomalous Jet Identification via Sequence Modeling
• Topological Obstructions to Autoencoding
• Unsupervised in-distribution anomaly detection of new physics through conditional density estimation
• The Efficacy of Event Isotropy as an Event Shape Observable
• Nonparametric semisupervised classification for signal detection in high energy physics
• Spheres To Jets: Tuning Event Shapes with 5d Simplified Models
• Unsupervised Event Classification with Graphs on Classical and Photonic Quantum Computers
• High-dimensional Anomaly Detection with Radiative Return in $e^{+}{e}^{-}$ Collisions
• Variational Autoencoders for Jet Simulation
• Uncovering hidden patterns in collider events with Bayesian probabilistic models
• An Exploration of Learnt Representations of W Jets

Cited by in corpus (20)

• Anomaly detection in high-energy physics using a quantum autoencoder
• Online-compatible Unsupervised Non-resonant Anomaly Detection
• Quantum Anomaly Detection for Collider Physics
• Deep Generative Models for Detector Signature Simulation: A Taxonomic Review
• Variational quantum one-class classifier
• What's Anomalous in LHC Jets?
• Anomaly Detection under Coordinate Transformations
• Nanosecond anomaly detection with decision trees and real-time application to exotic Higgs decays
• A semi-supervised approach to dark matter searches in direct detection data with machine learning
• Quantum Generative Adversarial Networks For Anomaly Detection In High Energy Physics
• Neural Embedding: Learning the Embedding of the Manifold of Physics Data
• Anomalies, Representations, and Self-Supervision
• Event-based anomaly detection for new physics searches at the LHC using machine learning
• Quantum support vector data description for anomaly detection
• Non-resonant Anomaly Detection with Background Extrapolation
• Unsupervised and lightly supervised learning in particle physics
• Simulation-based Anomaly Detection for Multileptons at the LHC
• Triggering Dark Showers with Conditional Dual Auto-Encoders
• Graph theory inspired anomaly detection at the LHC
• Creating Simple, Interpretable Anomaly Detectors for New Physics in Jet Substructure