Author:
Raine John Andrew,Klein Samuel,Sengupta Debajyoti,Golling Tobias
Abstract
We propose a new model independent technique for constructing background data templates for use in searches for new physics processes at the LHC. This method, called Curtains, uses invertible neural networks to parameterise the distribution of side band data as a function of the resonant observable. The network learns a transformation to map any data point from its value of the resonant observable to another chosen value. Using Curtains, a template for the background data in the signal window is constructed by mapping the data from the side-bands into the signal region. We perform anomaly detection using the Curtains background template to enhance the sensitivity to new physics in a bump hunt. We demonstrate its performance in a sliding window search across a wide range of mass values. Using the LHC Olympics dataset, we demonstrate that Curtains matches the performance of other leading approaches which aim to improve the sensitivity of bump hunts, can be trained on a much smaller range of the invariant mass, and is fully data driven.
Funder
Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung
Subject
Artificial Intelligence,Information Systems,Computer Science (miscellaneous)
Reference54 articles.
1. A generic anti-QCD jet tagger;Aguilar-Saavedra;J. High Energy Phys,2017
2. Simulation assisted likelihood-free anomaly detection;Andreassen;Phys. Rev. D,2020
3. ArdizzoneL.
KruseJ.
WirkertS.
RahnerD.
PellegriniE. W.
KlessenR. S.
Analyzing Inverse Problems With Invertible Neural Networks
4. Guided Image Generation With Conditional Invertible Neural Networks
5. The ATLAS experiment at the CERN large Hadron collider;J. Instrum,2008
Cited by
7 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献