SMuK 2023 – scientific programme

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

AKPIK: Arbeitskreis Physik, moderne Informationstechnologie und Künstliche Intelligenz

AKPIK 10: AI Topical Day – Computing II (joint session HK/AKPIK)

AKPIK 10.5: Talk

Thursday, March 23, 2023, 15:00–15:15, HSZ/0103

Optimization of the specific energy loss measurement for the upgraded ALICE TPC using machine learning — •Tuba Gündem for the ALICE Germany collaboration — Institut fuer Kernphysik, Frankfurt, Germany

The Time Projection Chamber (TPC) is the primary detector used in the ALICE experiment for tracking and particle identification (PID). PID is accomplished by reconstructing the momentum and the specific energy loss (dE/dx) of a particle. The dE/dx for a given track is calculated using a truncated mean on the charge signals associated to the track. The readout plane, on which the signals are measured, is radially subdivided into four regions with different pad sizes. Since the measured signals depend on the pad size, an optimization of the dE/dx calculation based on the pad size can be performed.

In this talk, a method for optimizing the dE/dx calculation using machine learning (ML) algorithms will be presented. By performing realistic simulations of the generated signals on the pads, various effects such as the different pad sizes and track geometry are modeled. These simulations are used as inputs for the training of the ML model and are investigated using RootInteractive.

Supported by BMBF and the Helmholtz Association.