Data Science
Part of the ATLAS High-Level trigger computing farm
Faculty
- James Brau
- Spencer Chang
- Tim Cohen
- Ben Farr
- Ray Frey
- Laura Jeanty
- Graham Kribs
- Stephanie Majewski
- Jake Searcy
- Dave Soper
- David Strom
- Eric Torrence
- Tien-Tien Yu
Data Science
As part of the University of Oregon’s Presidential Initiative in Data Science, IFS member Jake Searcy collaborates with researchers across campus to implement AI solutions for emerging research problems and is the Associate Director of AI for the UO’s Research Advanced Computing Services.
In Pine Mountain Observatory’s “Asteroid Light Curve” Project, students learn to operate PMO’s research-grade telescopes and dive into the realm of data analysis.
Collider-based particle physics and gravitational wave physics have been at the forefront of data science for decades. The world-wide web (first developed at CERN in 1989) is an early example of high energy physics (HEP) pushing the state of the art in computing.
As our research problems often involve sifting through enormous data sets (the current ATLAS data set comprises ~ 100 PB, or 100,000,000 GB of information), our analyses require using very large and internationally-distributed computing resources along with highly sophisticated analysis techniques to find exceedingly rare events in these immense data samples.
HEP has been at the forefront of developing grid computing as a scientific tool to solve the challenge of analyzing massive large hadron collider (LHC) data sets. ATLAS collaborators at Oregon run analysis jobs on a world-wide grid of interconnected computing facilities where the massive ATLAS data samples are stored.
HEP researchers have also been using machine learning techniques for many years to find very rare events buried in the mountains of data produced by our experiments. More recently, deep neural networks and other sophisticated machine learning techniques (largely developed by industry) are being increasingly applied to our analysis problems. The Oregon researchers are particularly interested in applying these sophisticated techniques in high-speed, FPGA-based custom electronics to improve the triggering of data at the LHC.