BACKPROPAGATION NEURAL NETWORKS - A TUTORIAL

Artificial neural networks have enjoyed explosive growth in the past ten years. An indication of the rate of growth of research in this area is the fact that, although the first research journal devoted exclusively to this subject was just introduced in 1987, there are now at least five refereed neural net research journals. These developments are being taken seriously by the semiconductor industry as well: in addition to a host of products developed by smaller firms, Intel, AT&T Bell Labs, Motorola and Hitachi have all introduced silicon implementations of neural network algorithms. Neural networks have a very broad scope of potential application, including many tasks central to chemical research and development. This tutorial begins with a short history of neural network research, and a review of chemical applications. The bulk, however, is devoted to providing a clear and detailed introduction to the theory behind backpropagation neural networks, along with a discussion of practical issues facing developers.