Autoencoders for unsupervised anomaly detection in high energy physics

Abstract

Autoencoders are widely used in machine learning applications, in particular for anomaly detection. Hence, they have been introduced in high energy physics as a promising tool for model-independent new physics searches. We scrutinize the usage of autoencoders for unsupervised anomaly detection based on reconstruction loss to show their capabilities, but also their limitations. As a particle physics benchmark scenario, we study the tagging of top jet images in a background of QCD jet images. Although we reproduce the positive results from the literature, we show that the standard autoencoder setup cannot be considered as a model-independent anomaly tagger by inverting the task: due to the sparsity and the specific structure of the jet images, the autoencoder fails to tag QCD jets if it is trained on top jets even in a semi-supervised setup. Since the same autoencoder architecture can be a good tagger for a specific example of an anomaly and a bad tagger for a different example, we suggest improved performance measures for the task of modelindependent anomaly detection. We also improve the capability of the autoencoder to learn non-trivial features of the jet images, such that it is able to achieve both top jet tagging and the inverse task of QCD jet tagging with the same setup. However, we want to stress that a truly model-independent and powerful autoencoder-based unsupervised jet tagger still needs to be developed.

Detecting New Physics as Novelty -- Complementarity Matters

“…Besides, it was also pointed out in Refs. [44] and [45] that the AE reconstruction error detects novelty only in one direction, as discussed in this paper. In relation to that, latent space tagging (a clustering-based method in our definition) was introduced to further improve the detection performance in Ref.…”

Cluster Scanning: a novel approach to resonance searches

“…Search for overdensity instead of outliers. Most anomaly search methods like Autoencoders [38] and SVDDs [88] rely on outlier detection, namely, identifying the data instances that lie in a region of very low probability density or outside the support of the "normal" distribution. Notably, while all normal events share similar characteristics and exhibit easily JHEP06(2024)163 recognisable trends, anomalous data, such as defects or fraud, can differ in numerous ways and are thus given a wide prior.…”

Machine Learning in the Search for New Fundamental Physics

“…Since the initial proposals of using autoencoders for anomaly detection [40][41][42], a number of improvements and modifications have been suggested. An important observation is that autoencoders can be biased by the relative complexity of anomalous and background data, potentially leading to outliers with a lower loss than the background [129][130][131]. As the latent space in VAEs [33,34] is optimised to follow a known distribution for backgrounds, it can also be used as anomaly score [132,133].…”