RanBox: anomaly detection in the copula space

Journal of High Energy Physics

Regular Article - Experimental Physics

Marija Mojsovska¹  Chiara Maccani¹  Giles C. Strong¹  Bruno Scarpa²  Martina Fumanelli²  Tommaso Dorigo³ 

[1] Dipartimento di Fisica e Astronomia “G.Galilei”, Università di Padova, Via F. Marzolo 8, 35131, Padova, Italy;Dipartimento di Scienze Statistiche, Università di Padova, Via C. Battisti 241, 35131, Padova, Italy;Istituto Nazionale di Fisica Nucleare, Sezione di Padova, Via F. Marzolo 8, 35131, Padova, Italy;Universal Scientific Education and Research Network (USERN), Tehran, Iran;

关键词: Hadron-Hadron Scattering;    Anomaly detection;    Unsupervised learning;   

DOI  :  10.1007/JHEP01(2023)008

 received in 2021-11-09, accepted in 2022-12-06,  发布年份 2022

来源: Springer

【 摘 要 】

The unsupervised search for overdense regions in high-dimensional feature spaces, where locally high population densities may be associated with anomalous contaminations to an otherwise more uniform population, is of relevance to applications ranging from fundamental research to industrial use cases. Motivated by the specific needs of searches for new phenomena in particle collisions, we propose a novel approach that targets signals of interest populating compact regions of the feature space. The method consists in a systematic scan of subspaces of a standardized copula of the feature space, where the minimum p-value of a hypothesis test of local uniformity is sought by greedy descent. We characterize the performance of the proposed algorithm and show its effectiveness in several experimental situations.

【 授权许可】

Unknown   
© The Author(s) 2023

【 图 表 】

40798_2022_490_Article_IEq46.gif