Abstract
The problem of data encoding for training back-propagation neural networks is well known. The basic principle is to avoid encrypting the underlying structure of the data. This is not easy in the real world, where we often receive data which has been processed by at least one previous user. The data may contain too many instances of some class, and too few instances of other classes. Then topology, and parameters settings designed. Finally, the network produces some results which need to be explained, or decrypted.
We present our experience and results on some satellite data augmented by a terrain model. The task was to predict the forest supra-type based on the available information. In this process we were forced to invent some methods to deal with very large amounts of erratically reliable data, and to produce meaningful predictions at the end.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Masters, T, Practical Neural Network Recipes in C, Academic Press, Boston, 1993.
Slade, P and Gedeon, TD “Bimodal Distribution Removal,” in Mira, J, Cabestany, J and Prieto, A, New Trends in Neural Computation, pp. 249–254, Springer Verlag, Lecture Notes in Computer Science, vol. 686, 1993.
Wong, PW and Gedeon, TD “The Error Sign Testing Method in Pattern Reduction,” International Journal of Systems Research and Information Science, (in press), 1994.
Gedeon, TD and Bowden, TG, “Heuristic Pattern Reduction II,” Proceedings ICCS, Invited Position Paper, pp. 3.43–3.45, Beijing, 1993.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1995 Springer-Verlag/Wien
About this paper
Cite this paper
Bustos, R.A., Gedeon, T.D. (1995). Decrypting Neural Network Data: A Gis Case Study. In: Artificial Neural Nets and Genetic Algorithms. Springer, Vienna. https://doi.org/10.1007/978-3-7091-7535-4_61
Download citation
DOI: https://doi.org/10.1007/978-3-7091-7535-4_61
Publisher Name: Springer, Vienna
Print ISBN: 978-3-211-82692-8
Online ISBN: 978-3-7091-7535-4
eBook Packages: Springer Book Archive