RanBox: anomaly detection in the copula space

Abstract

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.

Simulation-based anomaly detection for multileptons at the LHC

“…Recently, there have been many proposals for automating AD methods with machine learning [63][64][65][66][67][68][69][70][71][72][73][74][75][76][77][78][79][80][81][82] (see refs. [81][82][83][84] for overviews of the field).…”

What's anomalous in LHC jets?

“…In spite of this wealth of possible practical applications, the fundamental question still needs to be studied, namely what defines an anomaly search at the LHC? For large and stochas-tic datasets, the concept of outliers is difficult to define unambiguously, because any LHC jet or event configuration will occur with a finite probability, especially after we include detector imperfections [51][52][53][54][55]. In this situation, a simple, working definition of anomalous data is an event which lies in a low-density phase space region.…”

A normalized autoencoder for LHC triggers

“…While massive top jets stick out among low-mass QCD jets, we cannot expect a QCD jet to be special in a sample of top jets [55] since its simpler features will be interpolated by the neural network. A better-suited definition is based on low-probability regions in the background phase space distributions [56][57][58][59][60]. Here we can encode the phase space density, for instance, using cluster algorithms, VAEs, or a normalizing flow [15], but none of these methods are especially successful at identifying anomalous jets once the signal becomes more challenging than top jets.…”