Deep Learning Decouples Magnetic Gradient Inversion and Quantifies Three-Dimensional Subsurface Uncertainty
Researchers developed a physics-constrained neural network framework that analytically estimates magnetization vectors and maps subsurface targets with Bayesian uncertainty quantification.
