This review describes the IACT event reconstruction pipeline and the role of machine learning for classification and regression, highlighting timing features and ensemble methods as improvements over baseline approaches.
Improving the performance of the single-dish Cherenkov telescope MAGIC through the use of signal timing
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
The Cherenkov light flashes produced by Extensive Air Showers are very short in time. A high bandwidth and fast digitizing readout, therefore, can minimize the influence of the background from the light of the night sky, and improve the performance in Cherenkov telescopes. The time structure of the Cherenkov image can further be used in single-dish Cherenkov telescopes as an additional parameter to reduce the background from unwanted hadronic showers. A description of an analysis method which makes use of the time information and the subsequent improvement on the performance of the MAGIC telescope (especially after the upgrade with an ultra fast 2 GSamples/s digitization system in February 2007) will be presented. The use of timing information in the analysis of the new MAGIC data reduces the background by a factor two, which in turn results in an enhancement of about a factor 1.4 of the flux sensitivity to point-like sources, as tested on observations of the Crab Nebula.
fields
astro-ph.IM 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Machine Learning for Event Reconstruction in Imaging Atmospheric Cherenkov Telescopes
This review describes the IACT event reconstruction pipeline and the role of machine learning for classification and regression, highlighting timing features and ensemble methods as improvements over baseline approaches.