LEARNING AND GENERALIZATION CHARACTERISTICS OF THE RANDOM VECTOR FUNCTIONAL-LINK NET
Case Western Reserve University
Cleveland, OH 44106-7221
Of all the capabilities ascribable to neural-net computing, there is none more noteworthy than that of supervised learning. The generalized delta rule (GDR) net is preeminent in this task domain and is both widely known and widely used for learning functional mappings.
This learning capability can be used for assessment of the security or stability of systems; for monitoring the consistency of sets of sensor data; for predicting the future of systems; for estimating control actions necessary for achieving ...