AI techniques needed to advance particle-collider physics
DOI: 10.1063/PT.5.029414
Nature : Because huge amounts of data are being created by particle-collider experiments, researchers are seeking faster and more efficient ways to sift through the hundreds of millions of collisions that occur each second. At a CERN workshop last month, physicists and computer scientists met to discuss how advanced artificial intelligence (AI) techniques could be developed to spot patterns of both known particles and those yet to be discovered. That work would require a more sophisticated form of AI called deep learning, where the computer learns to recognize particle decay patterns on its own without human intervention. Workshop attendees also discussed how similar techniques could be used to find dark matter.
