A novel method for extracting knowledge from neural networks with evolving SQL queries

Date
2005-08-01
Authors
Gongora, Mario Augusto
Elizondo, David
Watson, Tim
Journal Title
Journal ISSN
ISSN
0302-9743
1611-3349
Volume Title
Publisher
Springer Verlag
Peer reviewed
Yes
Abstract
Description
The work reported in this paper provides the research community with a reliable and robust way to discern the way in which an NN based system makes its decision. It helps researchers tackle the problem of accepting the use of a system which is not understood unless its obscure internal knowledge is extracted and analysed for robustness and coherence.
Keywords
RAE 2008, UoA 23 Computer Science and Informatics, artificial neural networks, structured query language
Citation
Gongora, M.A., Watson, T., and Elizondo, D.A. (2005) A novel method for extracting knowledge from neural networks with evolving SQL queries. Lecture Notes in Computer Science, 3697, pp. 497-502.
Research Institute
Institute of Artificial Intelligence (IAI)