Jamie’s work focuses on handling the immense volumes of data produced by the LHCb experiment as it records collisions of protons up to 40 million times a second. To save collisions which contain interesting physics, this data must be processed and filtered in real time. His work focuses on this filtering, specifically on selection algorithms applied up to 1 million times a second. These algorithms use NNs to identify candidates containing distinct signatures of interest. He is responsible for developing novel techniques for selecting candidates, which requires a deep understanding of training and operating ML models, and the underlying physics case in which they are applied.
During a secondment to the ROOT analysis framework development team at CERN, he also worked on employing an emerging approach to statistical analysis, simulation-based inference (SBI), to LHCb physics analyses. In SBI, traditional statistical models are replaced by ML models, which are trained using simulations to learn the impact of parameters. This holds significant promise and could be trivially abstracted to other parameter estimation problems outside of a physics context.
I have always had a deep interest in physics, with many of my earliest memories being of the National Space Centre in Leicester. I was lucky enough to have excellent and encouraging physics teachers throughout school who had guided me along the path to studying for an integrated master’s degree in physics (MPhys) at the University of Manchester between 2018-2022.
It was whilst at Manchester that I became captivated by particle physics, obsessed with the puzzle of the imbalance between matter and antimatter in the universe. This obsession made the LHCb collaboration a very natural home, and through encouragement by my personal tutor, undergraduate laboratory work, a DAAD RISE internship at TU Dortmund and eventually a master’s project, I was able to become a part of the experiment.
