Anomaly Detection With Conditional Variational Autoencoders
This work addresses anomaly detection for particle physics experiments at CERN, offering a domain-specific solution that is incremental in adapting existing methods.
The paper tackled anomaly detection in hierarchically structured data, particularly for monitoring particle physics trigger systems, by proposing a conditional variational autoencoder with a novel loss function and metric, achieving superior performance on benchmarks and the real-world application.
Exploiting the rapid advances in probabilistic inference, in particular variational Bayes and variational autoencoders (VAEs), for anomaly detection (AD) tasks remains an open research question. Previous works argued that training VAE models only with inliers is insufficient and the framework should be significantly modified in order to discriminate the anomalous instances. In this work, we exploit the deep conditional variational autoencoder (CVAE) and we define an original loss function together with a metric that targets hierarchically structured data AD. Our motivating application is a real world problem: monitoring the trigger system which is a basic component of many particle physics experiments at the CERN Large Hadron Collider (LHC). In the experiments we show the superior performance of this method for classical machine learning (ML) benchmarks and for our application.