Timo Schellhaas ; Jens Sören Lange

2022
II. Institute of Physics Giessen

Abstract: Slow pions are classified based upon a pattern recognition algorithm using 9x9 pixel matrices in the PXD as input pattern. Background are electrons and positrons from QED processes. Support vector machines are, different from most of the other machine learning techniques, increasing the number of input dimensions, and following separating signal and background by hyperplanes. Values of 78% for correct identification were achieved.

Note: Presented on 28 03 2022
Note: BSc