Analysis & survey on fault tolerance in radial basis function networks

Conventional learning theory´s failure in training Neural Network to provide acceptable levels of generalization on the occurrences of fault in network has lead to the advent of Fault Tolerant Learning. Radial Basis Function networks are assumed to have in built Fault Tolerance capabilities. With this paper our attempt is to bring forth a detailed and time ordered survey of the literature available on Fault Tolerance in RBF networks. Methods, algorithms, measures for dealing with faults in RBF networks will be reported and analyzed. Future work along with directions is also presented.