Speakers
Description
AI/ML informed Symbolic Regression is the next stage of scientific modeling. We utilize a highly customizable symbolic regression package "PySR" to model the x and t dependence of the flavor isovector combination $H^{u-d}(x,t,ζ,Q^2)$ at ζ=0 and $Q^2$= 4 GeV$^2$. These PySR models were trained on GPD pseudodata provided by both Lattice QCD and phenomenological sources GGL, GK, and VGG. Symbolic convergence and consistency of the Lattice-Trained PySR models is demonstrated through the convergence of their Taylor expansion coefficients and their first moment in the forward limit, $A_{10}(t=0)$. In addition to PySR penalizing models with higher complexity and mean-squared error, we implement schemes that provide Force-Factorized and Semi-Reggeized PySR GPDs. We show that PySR disfavors a Force-Factorized model for non-factorizing GGL and GK sources, and that PySR Best Fit and Force-Factorized GPDs perform comparably well for the approximately factorizing VGG source.
*This work was supported by the DOE
