Recent Results from the France-Brazil Collaboration in the ATLAS experiment Framework: Machine Learning for HEP (Eduardo Simas)

par Eduardo Simas

lundi 22 janv. 2024, 14:00 → 16:00 Europe/Paris

amphi Charpak

A France-Brazil collaboration was established in 2018 to develop new signal processing and machine learning methods for high-energy calorimetry. Since then, researchers from both countries have been working together on many related topics within the ATLAS experiment collaboration. Some outcomes of this collaboration include a novel machine-learning energy calibration strategy for the ATLAS trigger system and the design of a hardware-based (FPGA) trigger for ATLAS phase II. A special effort was also devoted to designing and maintaining a general-purpose calorimeter simulation software (Lorenzetti Showers), which allows easy-of-use generation of different types of events, targeting the development and testing of novel algorithms and detector designs. This bilateral collaboration is expected to evolve towards the development of machine learning methods for dark matter searches as new funding was approved by CAPES-COFECUB for the next four years. For this, the Lorenzetti framework may be used together with generative AI for fast event simulation and reconstruction.
 

Zoom: https://cern.zoom.us/j/67907753293?pwd=QU9WdEpBbXB2VU5WSE5OZXQrRnN4UT09