Generative learning structures and processes for generalized connectionist networks

Massively parallel networks of relatively simple computing elements offer an attractive and versatile framework for exploring a variety of learning structures and processes for intelligent systems. This paper briefly summarizes some popular learning structures and processes used in such networks. It outlines a range of potentially more powerful alternatives for pattern-directed inductive learning in such systems. It motivates and develops a class of new learning algorithms for massively parallel networks of simple computing elements. We call this class of learning processes generative for they offer a set of mechanisms for constructive and adaptive determination of the network architecture-the number of processing elements and the connectivity among them-as a function of experience. Generative learning algorithms attempt to overcome some of the limitations of some approaches to learning in networks that rely on modification of weights on the links within an otherwise fixed network topology, for example, rather slow learning and the need for an a priori choice of network architecture. Several alternative designs as well as a range of control structures and processes that can be used to regulate the form and content of internal representations learned by such networks are examined. Empirical results from the study of some generative learning algorithms are briefly summarized, and several extensions and refinements of such algorithms and directions for future research are outlined.