Neural Network Modeling of Arbitrary Hysteresis Processes: Application to GO Ferromagnetic Steel