Münster 2017 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

T: Fachverband Teilchenphysik

T 21: Experimentelle Methoden 1 (Computing, Machine Learning, Statistik)

T 21.9: Vortrag

Montag, 27. März 2017, 18:45–19:00, JUR 253

Modern Machine Learning Methods in HEP — Raphael Friese, Günter Quast, Roger Wolf, and •Stefan Wunsch — Institut für Experimentelle Kernphysik, Karlsruhe, Germany

Modern machine learning methods such as deep neural networks are an active field of research in many scientific disciplines. Also the HEP community puts increasing effort in this emerging technology.

In particle physics, commonly used machine learning methods are boosted decision trees and shallow neural networks, which have proven their superior classification power over conventional cut based event selection in the last decade. Currently, deep learning shows again first signs of a significantly improved performance compared to these algorithms, which the HEP community aspires to exploit for its analyses.

This talk puts emphasis on the state-of-the-art usage of these modern machine learning methods and the application on event classification in particle physics.