Maximilian Nöthe

Improving the angular resolution of FACT using machine learning

FACT – The First G-APD Cherenkov Telescope is an Imaging Air Cherenkov Telescope operating at the Observatorio del Roque de los Muchachos on the Canary Island of La Palma, Spain. The main goals of FACT are to demonstrate the feasibility of Silicon Photo Multipliers comprised of Geiger Mode Avalanche Photo Diodes (G-APD) and to monitor the brightest known sources of gamma-ray emission as dense as possible. To improve the sensitivity of FACT for the detection of gamma-rays and thus improve the time until a possible change of state of the monitored sources can be detected, machine learning techniques are applied. In the past, background suppression and energy estimation were already done using random forests. Now also the directional reconstruction is applying methods of machine learning, which results in a great improvement compared to the existing, more simple method.