Machines to hunt ghost particles in big data
“KMI colloquium” 27th Feb (Wed) 15：00 –
Speaker:Kazuhiro TERAO (SLAC)
Title: Machines to hunt ghost particles in big data
Place: ES635 & ZOOM
Abstract: Explosive progress made in the area of Artificial Intelligence and Machine Learning (AI/ML) has been adopted in many areas of research and industries. Backed up with ever-evolving computing processors, AI/ML-enabled new ways to analyze data with improved quality as well as a new generation of techniques for the automation of workflows. In this talk, I will discuss my research using AI/ML to transform the short and long-baseline neutrino oscillation experiments in the U.S. with AI/ML techniques, including the Short Baseline Neutrino (SBN) program and Deep Underground Neutrino Experiment (DUNE). These experiments deploy Liquid Argon Time Projection Chamber (LArTPC) detectors that can capture a hi–resolution image of charged particles coming out of a neutrino interaction. While the detector technology shows strong promise, a quality data analysis technique is yet to be mastered. Traditional workflows to develop data reconstruction methods, detector calibrations, and physics analysis techniques require manual processes with lots of human interventions. Our research focus is to automate many of these workflows with acceleration using modern computing infrastructures and enabling new and quality ways to probe data for understanding the detector and unlocking the physics discovery potentials of these experiments.