Talk / Overview

Using AI for unsupervised anomaly detection offers an exciting opportunity for the model-independent discovery of new phenomena in fundamental physics. At the same time, the algorithms and technical challenges have a potential impact also for applications beyond particle physics. This talk will review the problem of anomaly detection in physics and introduce several machine-learning-based approaches to extract anomalies from data in an unsupervised way.

Talk / Speakers

Gregor Kasieczka

Professor, Universität Hamburg

AMLD / Global partners