| Sumario: | Part of a special section on geocomputation, which is defined as the process of applying computing technology to geographical problems. The articles in this special section demonstrate how things have moved on since the 1960s, when the size and complexity of the computations that could be handled were limited by the memory and speed of central processors. The ever-increasing speed of scalar computers and the availability of parallel computer architectures have changed this. The articles concentrate on methods, such as neural networks, genetic programming, simulation using cellular automata, simulation using random sampling of model parameters, fuzzy logic modeling, and geographically weighted regression, but they provide examples of their geographic application. The geographic problems to which they are applied include predicting spatial interaction between origins and destinations; projecting European Union regional populations; factors explaining spatial distribution of limiting long-term illness; the reconstruction of household types for small areas within cities; and the evolution of urban form over time.
|
|---|