Colloquium – Dylan Rankin (MIT)
Accelerating Discovery through Deep Learning
In recent years deep learning has become an increasingly common tool in high energy physics. Despite its prevalence, however, the most effective ways to use deep learning techniques are still unclear. In this talk I will discuss how advances in both hardware and software have allowed for faster and more efficient inference, and how these advances are enabling more sensitive detectors and powerful data analysis. I will also present new deep learning methods that are broadening the possibilities for searches and measurements at the Large Hadron Collider and beyond.
Host: Andrey Korytov