Combining outlier analysis algorithms to identify new physics at the LHC

Abstract

The lack of evidence for new physics at the Large Hadron Collider so far has prompted the development of model-independent search techniques. In this study, we compare the anomaly scores of a variety of anomaly detection techniques: an isolation forest, a Gaussian mixture model, a static autoencoder, and a β-variational autoencoder (VAE), where we define the reconstruction loss of the latter as a weighted combination of regression and classification terms. We apply these algorithms to the 4-vectors of simulated LHC data, but also investigate the performance when the non-VAE algorithms are applied to the latent space variables created by the VAE. In addition, we assess the performance when the anomaly scores of these algorithms are combined in various ways. Using super- symmetric benchmark points, we find that the logical AND combination of the anomaly scores yielded from algorithms trained in the latent space of the VAE is the most effective discriminator of all methods tested.

What's Anomalous in LHC Jets?

“…Based on these practical successes, ML-methods for anomaly detection at the LHC have generally received a lot of attention in the context of anomalous jets [10][11][12][13][14][15][16][17], anomalous events pointing to physics beyond the Standard Model [18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35], or enhancing established search strategies [36][37][38][39][40][41][42]. They include a first ATLAS analysis [43], experimental validation of some of the methods [44,45], quantum machine learning [46], applications to heavy-ion collisions [47], the DarkMachines challenge [48], and the LHC Olympics 2020 community challenge [49,50].…”

The Dark Machines Anomaly Score Challenge: Benchmark Data and Model Independent Event Classification for the Large Hadron Collider

“…21 First detailed in Ref. [118], this method involves training various anomaly detection methods within the latent space of a variational autoencoder, and then performing combinations of these anomaly scores to determine the optimal method. In the previously referenced paper we show that training an anomaly detection method on latent space representations of events dramatically improves the performance, and that combining these methods allows more information to be extracted.…”

The Dark Machines Anomaly Score Challenge: Benchmark Data and Model Independent Event Classification for the Large Hadron Collider

“…Research in ML has led to the development of new and enhanced anomaly detection methods that could be used and extended for applications employing LHC or astroparticle data. Examples of such outlier detection algorithms recently proposed for HEP include density-based methods [14], isolation forests, Gaussian mixture models [15], modelindependent searches with multi-layer perceptrons [16] , autoencoders [17][18][19][20], variational autoencoders [21,22], adversarially trained networks [23], or ML extended bump-hunting algorithms [24][25][26][27][28][29][30][31][32].…”