Deep Learning for Non-Perturbative Quantum Chromodynamics

Machine learning, particularly deep learning, is revolutionizing research across diverse disciplines, including physics. In this seminar, we explore the application of deep learning techniques to tackle challenges in non-perturbative Quantum Chromodynamics (QCD), one of the most complex areas in fundmental physics. I will present our preliminary explorations in this interdisciplinary field, focusing on: (i) identifying the equations of state for nuclear matter, (ii) developing a neural network-based quasi-particle model for QCD equations of state, (iii) extracting parton fragmentation functions, and (iv) determining heavy quark interaction potentials.

Fu-Peng Li s a Ph.D. candidate in Theoretical Physics at Central China Normal University(CCNU) with an expected graduation in June 2025. His research interests lie at the intersection of nuclear physics and machine learning, with a focus on auto-differentiation, physics-informed neural networks (PINNs) for inverse problems, and the application of machine learning to non-perturbative Quantum Chromodynamics (QCD).