معرفی
Dr Andrea Santamaria Garcia is a Lecturer in Artificial Intelligence for Particle Accelerators at the University of Liverpool and a member of the Cockcroft Institute. Her work bridges accelerator physics with machine learning, focusing on reinforcement learning and Bayesian optimization for real-time particle accelerator control.
- Current Research: Reinforcement learning in autonomous accelerators, Bayesian optimization, differentiable simulations
- Key Projects: RL4AA collaboration, Cheetah simulation code, microbunching instability control
- Education: PhD in Accelerator Physics (CERN), postdoctoral work on electron light sources
Her research explores the synergy between accelerator physics and machine learning, particularly through reinforcement learning and Bayesian optimization. Recent work focuses on zero-shot reinforcement learning, meta-algorithms, and deploying machine learning on hardware for low-latency beam control. She leads collaborations with KIT, DESY, and CERN, including the development of the first differentiable beam physics simulation code.
Scientific highlights include the 2024 Physical Review Accelerators and Beams Editors' Choice for her work on online reinforcement learning in accelerators. Her publications demonstrate cross-disciplinary impact in Scientific Reports and Physical Review Accelerators and Beams.
- PhD Supervision: Chenran Xu (KIT/DESY), Luca Scomparin (real-time RL platform)
- Teaching Philosophy: Emphasizes precision, ethics, and interdisciplinary collaboration
She collaborates with international institutions and co-founded the Reinforcement Learning for Autonomous Accelerators (RL4AA) initiative, which organizes annual workshops on cutting-edge control techniques.
