Elena Fol, CERN; Johann-Wolfgang-Goethe University
Machine Learning (ML) techniques are widely used in science and industry to discover relevant information and make predictions from data. Recently, the application of ML has grown also in accelerator physics, in particular in the domain of diagnostics and control.
Detection of faulty Beam Position Monitors using unsupervised learning already became a part of optics measurements at the LHC and is being successfully used to eliminate erroneous artifacts in the optics data.
Another subject of the current study is optics corrections using supervised learning. The optics correction results achieved with regression models convincingly demonstrate the great potential of this approach opening new opportunities for optics control in current and future accelerators. After a short introduction to ML concepts, both applications will be presented followed by the detailed discussion on the incorporated ML techniques.