article · Physical Review Letters
Unsupervised anomaly detection techniques can assist in the search for unknown physical phenomena within complex particle collision data. Using 140 inverse femtobarns of proton-proton collision data collected at 13 teraelectronvolts by the ATLAS detector at the Large Hadron Collider, collision events containing at least one electron or muon were analysed. An autoencoder neural network was trained directly on the data, with anomalous event regions identified through the reconstruction loss of the decoder. The analysis examined nine two-body invariant mass distributions consisting of a light jet or b-jet paired with an electron, muon, photon, or a second jet. Across these distributions, the observed data showed no significant deviation from standard background expectations. Consequently, upper limits were established for generic Gaussian resonance signals across various mass widths.
Exploring particle physics data without predefined signal models allows researchers to detect unexpected phenomena that traditional theory-driven searches might miss. Applying machine learning to sift through massive collision datasets enhances the ability to spot anomalies automatically. While no new physical particles were identified in this instance, establishing clear constraints helps refine future experimental searches and narrows the theoretical parameters for phenomena beyond the Standard Model.
The abstract does not indicate an application pathway, as it describes fundamental research in experimental particle physics.
AI-generated from the published abstract. Always read the original work before citing.
Searches for new resonances are performed using an unsupervised anomaly-detection technique. Events with at least one electron or muon are selected from 140 fb^{-1} of pp collisions at sqrt[s]=13 TeV recorded by ATLAS at the Large Hadron Collider. The approach involves training an autoencoder on data, and subsequently defining anomalous regions based on the reconstruction loss of the decoder. Studies focus on nine invariant mass spectra that contain pairs of objects consisting of one light jet or b jet and either one lepton (e,μ), photon, or second light jet or b jet in the anomalous regions. No significant deviations from the background hypotheses are observed. Limits on contributions from generic Gaussian signals with various widths of the resonance mass are obtained for nine invariant masses in the anomalous regions.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1103/physrevlett.132.081801
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
