Smart Control of Monte Carlo Simulations for Astroparticle Physics
share
Imaging Air Cherenkov Telescopes, such as VERITAS and MAGIC, rely on models for signal separation and energy estimation. These models are fit to simulated airshowers, each of which is associated with a single initial particle of known type and energy. Simulating an airshower from such primary particles requires the execution of fine-grained stochastic processes to emulate the behaviour of millions of secondary particles. In this work we explore the smart control of simulator runs to reduce the overall runtime of data generation whilst producing the most useful example set for model fitting. We give an overview of the approach and provide an evaluation using data from the widely used CORSIKA shower simulation.
READ FULL TEXT