Bonn 2020 – scientific programme

The DPG Spring Meeting in Bonn had to be cancelled! Read more ...

Parts | Days | Selection | Search | Updates | Downloads | Help

T: Fachverband Teilchenphysik

T 95: Experimental methods V

T 95.6: Talk

Friday, April 3, 2020, 12:15–12:30, H-HS VI

Classifying tau lepton decay modes using Deep Neural Networks at the ATLAS Experiment — •Hoang Nguyen, Klaus Desch, Philip Bechtle, Christian Grefe, Peter Wagner, Michael Hübner, and Lara Schildgen — Physialisches Institut, Uni Bonn, Deutschland

The tau lepton as the heaviest lepton in the Standard Model and plays an important role in many studies regarding Higgs physics or physics beyond the Standard Model. Of its decay modes, about two third occur hadronically.

The decay products of the tau lepton are difficult to distinguish from each other and other particles originating from jet and gluon interaction. A better knowledge of this could improve background suppression, help with studies of CP eigenstates and, furthermore, reconstruction accuracy will get better as well.

Latest studies indicate that the use of the predictive power of deep neural networks (DNN) yields better results than current likelihood or BDT based methods. In this presentation, a recurrent neural network with the aim to improve classification of tau decays is presented. Results obtained via this way are compared to https://arxiv.org/pdf/1512.05955.pdf.