This article presents a new learning algorithm for the construction and tra
ining of a RBF neural network. The algorithm is based on a global mechanism
of parameter learning using a maximum likelihood classification approach.
The resulting neurons in the RBF network partitions a multidimensional patt
ern space into a set of maximum-size hyper-ellipsoid subspaces in terms of
the statistical distributions of the training samples. An important feature
of the algorithm is that the learning process includes both the tasks of d
iscovering a suitable network structure and of determining the connection w
eights. The entire network and its parameters are thought of evolved gradua
lly in the learning process. (C) 1999 Elsevier Science Ltd. All rights rese
rved.