Abstract

This year marks the 10-year anniversary of the Higgs boson discovery. Since the start of data-taking at the LHC , it has been a long and complex journey, delivering further triumph to the Standard Model. Despite numerous searches for new physics, it remains elusive. With the future high energy physics experiments planned for decades ahead, we need to ask ourselves a question – where can we further innovate, and what might we have missed? In this talk, I will show the tremendous challenges that lie ahead of us at the High-Luminosity LHC , and I will argue that machine learning (ML) can help us solve them, while furthermore freeing resources for new ideas. With examples from state-of-the-art research, I will demonstrate how deep learning can improve, speed up, and optimise each stage of the data collection and analysis workflows at the LHC while extending the experimental sensitivity. Finally, by showing the physics impact of the ML solutions, I hope to convince you that machine learning is not only the past and present of particle physics, but it has to be the future as well.