Workshop on machine learning applications for particle accelerators
Dr Ing. Gianluca Valentino (Faculty of ICT) was one of the co-organisers of a workshop on machine learning applications for particle accelerators, which was held between 27 February and 2 March 2018 at the SLAC National Accelerator Laboratory in Menlo Park, CA, USA (agenda).
The workshop was held under the auspices of the International Committee for Future Accelerators (ICFA), and gathered together over 60 scientists from a wide variety of laboratories and universities around the world, including CERN, Fermilab, SLAC, DESY, PSI, Cornell University, BNL, ETH Zurich and KIT.
The purpose of the workshop was to collect the community’s current understanding of relevant state-of-the-art machine learning techniques, as well as their application to tuning, control, prognostics, anomaly detection and simulation/modeling in particle accelerators, which is a growing area of research. Dr Ing. Valentino delivered a tutorial on unsupervised learning techniques, and presented ongoing research which he has conducted on using machine learning to automatically classify between beam loss planes using data from Beam Loss Monitors at the Large Hadron Collider in collaboration with CERN. This work has also been accepted for publication at the International Particle Accelerator Conference which will be held in May 2018 in Vancouver, Canada.
Several attendees at the workshop, including Dr Ing. Valentino, are presently drafting a whitepaper to illustrate the current opportunities for machine learning techniques to optimize the design, operation and performance of particle accelerators.