Weakly supervised anomaly detection with event-level variables
arXiv:2504.13249 · doi:10.1103/rk29-518p
Abstract
We introduce a new topology for weakly supervised anomaly detection searches, di-object plus~X. In this topology, one looks for a resonance decaying to two standard model particles produced in association with other anomalous event activity (X). This additional activity is used for classification. We demonstrate how anomaly detection techniques which have been developed for di-jet searches focusing on jet substructure anomalies can be applied to event-level anomaly detection in this topology. To robustly capture event-level features of multi-particle kinematics, we employ new physically motivated variables derived from the geometric structure of a collision's phase space manifold. As a proof of concept, we explore the application of this approach to several benchmark signals in the di- and di- plus~X final states. We demonstrate that our anomaly detection approach can reach discovery-level significances for signals that would be missed in a conventional bump-hunt approach.
15 Pages, 8 figures, Submitted to Physics Review D
References in corpus (56)
- The automated computation of tree-level and next-to-leading order differential cross sections, and their matching to parton shower simulations
- An Introduction to PYTHIA 8.2
- FastJet user manual
- DELPHES 3, A modular framework for fast simulation of a generic collider experiment
- Parton distributions for the LHC Run II
- Comparative study of various algorithms for the merging of parton showers and matrix elements in hadronic collisions
- The Next-to-Minimal Supersymmetric Standard Model
- Measurement of an Excess of B -> D(*) Tau Nu Decays and Implications for Charged Higgs Bosons
- Automatic spin-entangled decays of heavy resonances in Monte Carlo simulations
- Measurement of the branching ratio of relative to decays with hadronic tagging at Belle
- Measurement of the ratio of branching fractions
- A first unbiased global NLO determination of parton distributions and their uncertainties
- SUSY Les Houches Accord 2
- Search for high-mass dilepton resonances in pp collisions at sqrt(s) = 8 TeV with the ATLAS detector
- Averages of -hadron, -hadron, and -lepton properties as of 2021
- Test of Lepton Flavor Universality by the measurement of the branching fraction using three-prong decays
- Search for new high-mass phenomena in the dilepton final state using 36 fb of proton-proton collision data at = 13 TeV with the ATLAS detector
- Classification without labels: Learning from mixed samples in high energy physics
- Searching for New Physics with Deep Autoencoders
- Anomaly Detection for Resonant New Physics with Machine Learning
- Automatic evaluation of UV and R2 terms for beyond the Standard Model Lagrangians: a proof-of-principle
- Measurement of and with a semileptonic tagging method
- Measurement of the lepton polarization and in the decay with one-prong hadronic decays at Belle
- Search for heavy Higgs bosons decaying into two tau leptons with the ATLAS detector using collisions at TeV
- Extending the Bump Hunt with Machine Learning
- Performance of reconstruction and identification of leptons decaying to hadrons and in pp collisions at 13 TeV
- QCD or What?
- Simulation Assisted Likelihood-free Anomaly Detection
- Search for high-mass resonances in dilepton final states in proton-proton collisions at 13 TeV
- Dijet resonance search with weak supervision using TeV collisions in the ATLAS detector
- The Dark Machines Anomaly Score Challenge: Benchmark Data and Model Independent Event Classification for the Large Hadron Collider
- High-pT Signatures in Vector-Leptoquark Models
- Classifying Anomalies THrough Outer Density Estimation (CATHODE)
- Benchmark scenarios for the NMSSM
- Tag N' Train: A Technique to Train Improved Classifiers on Unlabeled Data
- Searches for additional Higgs bosons and for vector leptoquarks in final states in proton-proton collisions at = 13 TeV
- Test of lepton flavour universality using decays with hadronic channels
- Machine Learning for Anomaly Detection in Particle Physics
- Identification of hadronic tau lepton decays using a deep neural network
- Resonant anomaly detection without background sculpting
- Search for long-lived particles in final states with displaced dimuon vertices in collisions at 13 TeV with the ATLAS detector
- Comparing Weak- and Unsupervised Methods for Resonant Anomaly Detection
- FETA: Flow-Enhanced Transportation for Anomaly Detection
- Search for new phenomena in two-body invariant mass distributions using unsupervised machine learning for anomaly detection at TeV with the ATLAS detector
- Anomaly detection search for new resonances decaying into a Higgs boson and a generic new particle in hadronic final states using TeV collisions with the ATLAS detector
- Search for new phenomena in collisions in final states with tau leptons, -jets, and missing transverse momentum with the ATLAS detector
- Review of Semileptonic Anomalies
- Model-agnostic search for dijet resonances with anomalous jet substructure in proton-proton collisions at = 13 TeV
- Searching in CMS Open Data for Dimuon Resonances with Substantial Transverse Momentum
- Covariantizing Phase Space
- Search for pair production of third-generation leptoquarks decaying into a bottom quark and a -lepton with the ATLAS detector
- Combining Resonant and Tail-based Anomaly Detection
- Search for a third-generation leptoquark coupled to a lepton and a b quark through single, pair, and nonresonant production in proton-proton collisions at = 13 TeV
- Mutually elusive: vectorlike antileptons and leptoquarks
- Search for a high-mass dimuon resonance produced in association with b quark jets at = 13 TeV
- Study of DiMuon Rare Beauty Decays with ATLAS and CMS