PhD Colloquium: Borut Kersevan

This seminar presents the diverse use of Machine Learning tools and approaches in High Energy Physics. It focuses on the practices of the experiments at the Large Hadron Collider (LHC) at CERN. The LHC experiments use Machine Learning in a wide range of applications. These range from straightforward procedures of trying to differentiate the new physics and known processes in data analysis using e.g. deep neural networks, to using Machine Learning to replace traditional Monte Carlo simulation of physics processes.
State-of-the-art attempts comprise using programmable FPGA chips to implement very fast Machine Learning tools in detector operations, exploring the use of Machine Learning algorithms on Quantum Computers, employ Artificial Intelligence approaches to design the new generations of experiments, solve theoretical equations, etc...
Special emphasis will be given to the implementation of the transfer of latest commercial approaches, such as generative modelling, into scientific procedures with advantages they bring as well as associated caveats. Finally, a speaker’s vision of the future of Machine Learning in HEP will be given.