Bereiche | Tage | Auswahl | Suche | Downloads | Hilfe

T: Fachverband Teilchenphysik

T 96: Kosmische Strahlung 4

T 96.7: Vortrag

Dienstag, 10. März 2009, 18:15–18:30, M105

Composition studies by kNN procedure with KASCADE-Grande data — •Fabiana Cossavella — Universität Karlsruhe, Institut für Experimentelle Kernphysik, 76021 Karlsruhe

KASCADE-Grande, located at Forschungszentrum Karlsruhe, is a multi-detector experiment for the measurement of extensive air showers induced by primary cosmic rays in the energy range of 10¹⁴ − 10¹⁸ eV.

The “k-Nearest Neighbours” (kNN) method is a classification procedure applied for a preliminary study of the cosmic ray composition in this energy range. Simulations of different primary particles are used as reference samples. In order to find for each real event the k Nearest Neighbours in the reference sample, the Mahalanobis distance in the space defined by the choosen observables is calculated. The probability of the event to be part of one of the simulated primary groups is the percentage of the k neighbours belonging to it.

An attempt to separate light from heavy primaries in the energy range of KASCADE-Grande is here reported.