Fundamentals in Neurocomputing

Neurocomputing - inspired from neuroscience - provides the potential of an alternative
information processing paradigm that involves large interconnected networks of relatively
simple and typically non-linear processing elements, so-called (artificial) neural networks.
There has been a recent resurgence in the field of neural networks, caused by new net
topologies and algorithms, and the belief that massive parallelism is essential for high
peiformance in several research areas, especially in pattern recognition. This contribution
provides a brief introduction to some basic features of neural networks by defining a neural
network, reflecting current thinking about the processing that should be peiformed at each
processing element of a neural network, discussing the general categories of training that are
commonly used to adjust a neural network's weight vector, and finally by characterizing the
backpropagation neural networ:k which is one of the most important historical developments
in neurocomputing.- The contribution concludes with pointing to some hot topics for future
research. It is hoped that this contribution will stimulate the study of neural networks in
quantitative geography and regional science. (author's abstract)