Two applications of neural networks on spatial data
This paper provides some insight into connectionist methods which have been used to address two different tasks involving spatial data sets. The first application involves supervised classification of remotely sensed imagery. It demonstrates the potential of neural networks for performing complex classification tasks. An artificial neural network based on the multilayer perceptron model was used to classify two-date multispectral SPOT high resolution visible imagery. The second application uses the Kohonen Maps connectionist model to perform unsupervised data analysis on European regions described by socio-economic variables. The data set under study was supplied by the Statistical Office of the European Community (EUROSTAT).
Bibliographic Reference: Paper presented: Workshop on New Tools for Spatial Analysis, Lisboa (PT), November 16-21, 1993
Availability: Available from (1) as Paper EN 37957 ORA
Record Number: 199311658 / Last updated on: 1994-11-28
Original language: en
Available languages: en