A learning algorithm for connectionist concept classes

A connectionist learning algorithm, the BRD (bounded, random, distributed) algorithm, is defined and formally analyzed within the framework of computational learning theory. From a neural network viewpoint this framework gives clear definitions to such commonly used terms as ´generalization´ and ´scaling up´ and addresses such issues as what class of functions is being learned, how many training examples should be used, how many iterations are required, and with what certainty can one be assured of learning a good model? From a computational learning theory perspective, a new class of connectionist concepts is shown to be polynomially learnable using the BRD algorithm. Since a variant of the BRD algorithm is in current use for tasks such as pattern recognition, this makes it one of the few learning algorithms shown to be polynomial within the computational learning theory framework that is close to an ´industrial strength´ algorithm.<>