Choose timezone
Your profile timezone:
Powered by Indico
v3.3.3
The Higgs boson was famously co-discovered in 2012 by the ATLAS experiment at CERN. Since then, the collaboration has studied in detail the various properties of this particle, long heralded as "the last missing piece of the Standard Model". In recent years, machine learning techniques have found a large range of applications in high energy particle physics. For ATLAS analyses, this has translated into faster and more accurate simulation and reconstruction algorithms, improved particle taggers and new discriminators to isolate rare processes from large quantities of background events. And yet, even though our understanding of the Standard Model has been greatly refined since the discovery of the Higgs boson, no new elementary particle (the clearest possible sign of physics Beyond the Standard Model) has been observed so far. With extensive but conventional search programmes only leading to null results, the key to a new discovery might lie with unsupervised machine learning: a model-independent way of sifting through the large and complex datasets extracted by the ATLAS experiment and pointing out any potential anomalies.
In this talk, I will describe the ATLAS detector and recent developments in machine learning for collider physics; no prior knowledge about either topic is required.