Undergraduate Research at Jefferson Lab

AI in QCD Global Analysis

Student: Joshua Miller

School: Lebanon Valley College

Mentored By: Nobuo Sato

Protons and neutrons are composed of three quarks each. The proton has two up and a down quark while the neutron has two down and an up quark. These quarks are connected together by gluons. These quarks have parton distribution functions (PDFs) which are dependent on parameters in the form of α and β. These PDFs are used in calculating the cross-sections of collisions with a proton and a neutron. These cross sections are σ^p_red = 4u(α,x) + d(β,x) and σⁿ_red = u(α,x) + 4d(β,x). The goal of the experiment is to use machine learning and create a Generative Adversarial Neural Network (GAN) that is able to produce these parameters and calculate cross-section values. The GAN was able to successfully produce parameter values that can calculate cross-section values similar to true calculations. From this, we are able to better understand the structure of protons and neutrons due to them being related to quantum correlation functions, which are extracted from this GAN. This research will then be used in a global analysis of protons and neutrons which will be used in later research for the Department of Energy. From this research, there was a lot of coding in Python which taught valuable lessons. You will not know everything from the beginning, and it is okay if your research is not correct the first time around since you are still conducting research. Professional development was also made throughout the past weeks. One professional development that was taught was video conference etiquette. With COVID-19, it was important to learn the proper ways to interact with video calls. Another professional development was learning time management. This is important because meetings had to be scheduled between multiple people when there is only one time a week when everyone can meet. Time management is also important because there are deadlines that must be met.

[Watch the presentation on YouTube]