JAI Seminar: Machine Learning Techniques for Particle Accelerators

On Thursday 20th February, new University of Oxford researcher Dr. Hector Garcia-Morales, presented a seminar at the JAI on using machine learning techniques for particle accelerators.

Particle accelerators can be rather complex scientific systems to operate and understand. The large amounts of data and the non-linear relationship between the different accelerator parameters makes the optimisation of their operation performance very challenging. Today, Machine Learning techniques are mature enough to be applied to accelerators. These tools are potentially very valuable to meet the operational demands of current and future accelerators to optimise their performance, automatise procedures and detect faults. In the seminar, Hector explored the potential of using such techniques and discussed the latest advances in the field with some applications to operating accelerators.

Dr. Garcia-Morales completed a PhD in accelerator physics in 2015 at the Universitat Politècnica de Catalunya. Since then he has been a member of the JAI, first working as a postdoctoral researcher at Royal Holloway, University of London, until 2019 when he took up his present role at the University of Oxford. Spending several years working at CERN Hector's research projects have included the ILC, CLIC and HL-LHC as well as several others. Throughout his career Hector has been heavily involved with science communication, writing for several publications and creating videos for his YouTube channel.