AN ACTIVE LEARNING APPROACH FOR RADIAL BASIS FUNCTION NEURAL NETWORKS

This paper presents a new Active Learning algorithm to train Radial BasisFunction (RBF) Artificial Neural Networks (ANN) for model reduction problems. The newapproach is based on the assumption that the unobserved training data y at input x, lieswithin a setF (x) ={ y: f (x) ≤ y ≤ f (x)} where F(x) is known from experience or pastsimulations. The new approach finds the location of the new sample such that the worst caseerror between the output of the resulting RBF ANN and the bounds of the unknown data asspecified by F(x) is minimized. This paper illustrates the new approach for the case when x Є R1. It was found that it is possible to find a good location for the new data sample byusing the suggested approach in certain cases. A comparative study was also done indicatingthat the new experiment design approach is a good complement to the existing ones such ascross validation design and maximum minimum design.