Machine learning in the search for new fundamental physics

Compelling experimental evidence suggests the existence of new physics beyond the well-established and tested standard model of particle physics. Various current and upcoming experiments are searching for signatures of new physics. Despite the variety of approaches and theoretical models tested in these experiments, what they all have in common is the very large volume of complex data that they produce. This data challenge calls for powerful statistical methods. Machine learning has been in use in high-energy particle physics for well over a decade, but the rise of deep learning in the early 2010s has yielded a qualitative shift in terms of the scope and ambition of research. These modern machine learning developments are the focus of the present Review, which discusses methods and applications for new physics searches in the context of terrestrial high-energy physics experiments, including the Large Hadron Collider, rare event searches and neutrino experiments. Owing to the growing volumes of data from high-energy physics experiments, modern deep learning methods are playing an increasingly important role in all aspects of data taking and analysis. This Review provides an overview of key developments, with a focus on the search for physics beyond the standard model. There have been large and sustained developments of deep learning in high-energy physics over the past several years. Supervised machine learning methods are widely used to identify known particles and to design targeted searches for specific theories of new physics. Less-than-supervised machine learning methods are used to carry out searches that depend less on a specific signal model. Experiments such as those at the Large Hadron Collider, neutrino detectors and rare event searches for dark matter, despite having different technical requirements, also share similarities, and there is ground for cooperation to develop machine learning methods. Combining physics and new ideas from statistical learning will be crucial to analysing the large volumes of data to potentially uncover the fundamental structure of nature. There have been large and sustained developments of deep learning in high-energy physics over the past several years. Supervised machine learning methods are widely used to identify known particles and to design targeted searches for specific theories of new physics. Less-than-supervised machine learning methods are used to carry out searches that depend less on a specific signal model. Experiments such as those at the Large Hadron Collider, neutrino detectors and rare event searches for dark matter, despite having different technical requirements, also share similarities, and there is ground for cooperation to develop machine learning methods. Combining physics and new ideas from statistical learning will be crucial to analysing the large volumes of data to potentially uncover the fundamental structure of nature.