Speaker
Description
We investigate the thermodynamic and transport properties of the quark-gluon plasma at finite baryon chemical potential within a deep-learning-assisted quasi-particle model. The temperature and baryon chemical potential-dependent effective masses of partons are estimated using residual neural networks trained on lattice QCD results for the equation of state. Further, mass-regularization constraints are imposed to preserve a consistent quasi-particle mass hierarchy. The trained framework is subsequently employed to evaluate the thermodynamic observables and transport coefficients of the deconfined medium. Our model predictions for the speed of sound and specific heat are in good agreement with lattice estimations. Moreover, we study the ratios of the shear and bulk viscosities to the entropy density, as well as the scaled electrical and thermal conductivities. Our results are qualitatively comparable to available lattice calculations and other phenomenological model predictions. These results demonstrate that a deep-learning-assisted quasiparticle framework provides a flexible and efficient approach to extending QCD thermodynamics and transport properties into the finite-baryon-chemical-potential regime.