Cosmo-PINN: A Physics-Informed Neural Network for Cosmological Reconstruction
Cosmo-PINN uses a Physics-Informed Neural Network to reconstruct dark energy's equation of state parameter wDE(z) from late-time cosmological data like Supernovae and BAO. By embedding physical laws directly into the training loss function, it ensures the reconstructed model is physically consistent, unlike purely data-driven methods which can yield unphysical results.