Abstract:Based on lattice QCD results of equation of state (EOS) and baryon number susceptibility at zero baryon chemical potential, and supplemented by machine learning techniques, we construct the analytic form of the holographic black hole metric in the Einstein-Maxwell-Dilaton (EMD) framework for pure gluon, 2-flavor, and 2+1-flavor systems, respectively. The dilaton potentials solved from Einstein equations are in good agreement with the extended non-conformal DeWolfe-Gubser-Rosen (DGR) type dilaton potentials fixed by lattice QCD EOS, which indicates the robustness of the EMD framework. The predicted critical endpoint (CEP) in the 2+1-flavor system is located at $(T^c$=0.094GeV, $\mu^c_B$=0.74GeV), which is close to the results from the realistic Polyakov-Nambu-Jona-Lasinio(PNJL) model, the functional renormalization group, and the holographic model with extended DeWolfe-Gubser-Rosen dilaton potential.

What problem does this paper attempt to address?

This paper attempts to construct an analytic form of holographic black - hole metric by combining the results of lattice quantum chromodynamics (QCD) and machine - learning techniques, in order to describe the equation of state (EOS) of pure - gluon, 2 - flavor and 2 + 1 - flavor systems in the Einstein - Maxwell - Dirac (EMD) framework. Specifically, the paper aims to: 1. **Utilize machine - learning techniques**: Extract the equation of state and baryon number susceptibility at zero baryon chemical potential from lattice QCD results, and then determine the black - hole metric in the EMD model. 2. **Construct an analytic form of black - hole metric**: In the EMD framework, compare the Dirac potential obtained by solving the Einstein equation with the extended non - conformal DeWolfe - Gubser - Rosen (DGR) - type Dirac potential fixed by the lattice QCD EOS to verify the robustness of the EMD framework. 3. **Predict the location of the critical end point (CEP)**: Especially in the 2 + 1 - flavor system, predict the location of the critical end point and compare it with the results of the actual Polyakov - Nambu - Jona - Lasinio (PNJL) model, the functional renormalization group (FRG), and the holographic model with the extended DGR Dirac potential. ### Main problems and goals - **Problems to be solved**: How to accurately describe the equation of state of QCD matter and predict the location of the critical end point using machine - learning and holographic duality methods under finite temperature and density conditions. - **Goals**: Verify the validity of the EMD framework and predict the location of the critical end point in different systems by constructing an analytic form of holographic black - hole metric. ### Methods and techniques - **Data sources**: Use the results of lattice QCD, including the equation of state and baryon number susceptibility of pure - gluon, 2 - flavor and 2 + 1 - flavor systems. - **Machine - learning model**: Adopt a deep neural network for regression analysis and optimize parameters to fit the lattice QCD data. - **Model verification**: Verify the accuracy and reliability of the model by comparing the machine - learning results with the lattice QCD results. ### Results - **Construction of analytic metric**: Successfully constructed holographic black - hole metrics for pure - gluon, 2 - flavor and 2 + 1 - flavor systems. - **Prediction of critical end point**: Predicted the location of the critical end point in the 2 + 1 - flavor system as $(T_c = 0.094 \text{GeV}, \mu_c^B = 0.74 \text{GeV})$ and compared it with the results of other non - perturbative models, finding it to be very close. ### Conclusions - **Effectiveness of the model**: The holographic model constructed by machine - learning shows good accuracy in describing the equation of state of QCD matter, especially for the prediction of the critical end point. - **Future prospects**: It is hoped that this method can help to gain a deeper understanding of the hadron spectrum in the strong - interaction field and further integrate more information in future work to construct more realistic holographic QCD models.

Machine learning holographic black hole from lattice QCD equation of state

Machine learning holographic black hole from lattice QCD equation of state

QCD Phase Diagram at finite Magnetic Field and Chemical Potential: A Holographic Approach Using Machine Learning

Flavor dependent Critical endpoint from holographic QCD through machine learning

Revisiting holographic model for thermal and dense QCD with a critical point

Hot QCD Phase Diagram From Holographic Einstein-Maxwell-Dilaton Models

Deriving dilaton potential in improved holographic QCD from chiral condensate

Towards a warm holographic equation of state by an Einstein-Maxwell-dilaton model

Spectra of glueballs and oddballs and the equation of state from holographic QCD

Holographic black hole engineering at finite baryon chemical potential

Holographic Q-lattices

Dynamical Holographic QCD Model: Resembling Renormalization Group from Ultraviolet to Infrared

Analytic study on chiral phase transition in holographic QCD

Phase structure and critical phenomena in two-flavor QCD by holography

Holographic reconstruction of black hole spacetime: machine learning and entanglement entropy

Simulating black hole quantum dynamics on an optical lattice using the complex Sachdev-Ye-Kitaev model

Holographic entanglement properties in the QCD phase diagram from EMD models

Phase structure and critical phenomena in 2-flavor QCD by holography

Learning the black hole metric from holographic conductivity

Machine Learning Holographic Mapping by Neural Network Renormalization Group

Deep-learning Quasi-Particle Masses from QCD Equation of State