Unsupervised and Generative AI for High Energy Physics and Beyond

• Kevin Pedro (Fermi National Accelerator Lab. (US))

The next generation of high energy particle physics experiments and astrophysical observatories will provide an unprecedented flood of data. Cutting-edge AI-based density estimation techniques are essential components of any program to maximize the discovery potential of this data. Unsupervised learning can improve interpretability and robustness by ensuring generalization, as well as enabling anomaly detection to search for new physics in unexpected areas. Generative learning can provide faster, higher quality simulations to match increasing data volumes and precision requirements. These new algorithms can be deployed on coprocessors such as GPUs and neuromorphic chips to accelerate data processing far beyond conventional computing. We will discuss promising current and future applications to collider physics and multimessenger astronomy. You received this email because you are subscribed to the high energy physics experimentalists mailing list (HEPEX) mailing list. If you would like to unsubscribe from this list, simply send an email to listserv@listserv.umd.edu with the message signoff HEPEX in the body.