Speaker
Auralee Edelen
(SLAC)
Description
Diagnostic methods that are enhanced with machine learning are improving the speed and detail with which beam behavior can be characterized on-the-fly in real accelerator systems. Detailed characterization can in turn improve both high-precision modeling of accelerator systems and high-precision optimization/control for high brightness beams. This talk will outline the state-of-the-art in machine learning enhanced diagnostics for accelerators, ranging from fast data-driven approaches for shot-to-shot prediction to methods that tightly couple machine learning and physics simulations for unprecedented fidelity in beam phase space reconstruction.
Primary author
Auralee Edelen
(SLAC)
