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

HK: Fachverband Physik der Hadronen und Kerne

HK 13: Instrumentation V

HK 13.5: Talk

Monday, February 26, 2018, 17:45–18:00, HZO 80

Fuzzy Bayes Tracking – Experimental performance — •Philipp Napiralla¹, Herbert Egger², Philipp R. John¹, Norbert Pietralla¹, Michael Reese¹, and Christian Stahl¹ — ¹Institut für Kernphysik, TU Darmstadt — ²AG Numerik und wissenschaftliches Rechnen, TU Darmstadt

The Advanced GAmma Tracking Array (AGATA) is a new type of γ-spectrometer using position resolution via Pulse Shape Analysis to allow for a high energy resolution in addition to high efficiency. Due to its Germanium shell without any Compton shielding, γ-ray tracking algorithms are needed. In contrast to existing tracking algorithms that are based on a “Figure of Merit” approach, the so-called Bayes-Tracking algorithm uses conditional probability densities and Bayes’ Theorem. As an improvement of the Bayes-Tracking algorithm, the presented Fuzzy Bayes Tracking (FBT) introduces fuzzy logic and machine learning into the framework. FBT’s performance on experimental source data is shown and compared to existing tracking algorithms. An outlook on possible experimental applications of the Fuzzy Bayes Tracking is given.

Supported by BMBF 05P15RDFN1 and 05P15RDFN9.