Publication | Closed Access

A dimension-reduced variational approach for solving physics-based inverse problems using generative adversarial network priors and normalizing flows

Numerical AnalysisEngineeringMachine LearningVariational AnalysisPhysics-based Inverse ProblemsPde-constrained OptimizationGenerative Adversarial NetworkDimension-reduced Variational ApproachGenerative ModelInverse ProblemsNormalizing Flows