Abstract: Traditional beam tomography techniques require hundreds of samples for a high-fidelity reconstruction. The performance of heavy ion accelerators, such as FRIB and ATLAS, may suffer from the drift of the accelerator, resulting in the need for efficient reconstruction using fewer samples before beam distribution changes significantly. This work uses a machine-learning approach to reconstruct the 2D or 4D phase space directly from 1D measurements.
Bio: Anthony was born and raised in southern California. He went to the University of California San Diego for his physics undergrad, doing some plasma simulation research. Now, he is working under Professor Yue Hao at Michigan State University as a 4th-year graduate student in Accelerator Physics. His current field of study uses machine learning in accelerator optimization and beam tomography.
