Miaoyuan Liu
[introductory/intermediate] Edge of the Future: AI in Real Time Systems of Scientific Instruments
Summary
Real-time AI inference on targetted systems or heteregeous systems have potential to revolutionize trigger and data acquisition sytems of scientific instruments. In this course, I will introduce basic elements for real-time ML inference for science with a few examples on the Compact Muon Collider experiment.
Syllabus
References
https://indico.cern.ch/event/737461/contributions/4040692/
Pre-requisites
Basic knowledge of fully connected neural networks, convolutional neural networks and graph neural networks.
Short bio
Dr. Liu is an Assistant professor at Purdue University working on the CMS experiment on the LHC. My research focuses on using machine learning to advance scientific discovery potentials. Currently, my group focuses on 1) Nanosecond ultra-low latency Geometric Deep Learning (GDL) for real-time systems such as the CMS hardware trigger, 2) Building ML methods to study the interpretability and generalizability of GDL, 3) Efficient processing of large data sets with heterogeneous computing resources.
