Fuzzy-neunet: A non standard neural network
Most of today's connectionist networks hold the information in the weights of the connections, the synapses (see Error Back Propagation, Hopfield-Net, Neocognitron). In contrast to these models NEUNET is a fully self organizing network. Its information is represented only in its overall structure, which is adopted dynamically through new ‘experiences’ and a special type of persistent activation-states (the so-called stamps) of the units.
The goal of this paper is to give an overview of the NEUNET-algorithms as well as of its theoretical background. The main part is dedicated to a new probability-based approach which does significantly improve the capabilities of NEUNET. Some characteristic examples are given for illustrating applications in pattern recognition with autoassociative recall. In addition to a presentation of the current state of development of NEUNET a description of prospects of future work is given.
Unable to display preview. Download preview PDF.
- [And88]ANDLINGER,P.: Serielle Assoziation in neuronennetzähnlichen Netzwerken — in: Proc. of the 33rd International Scientific Colloquium, Ilmenau Germany (1988)Google Scholar
- [Fuk80]FUKUSHIMA, K.: Neocognitron: A Self-organizing Neural Network Model for a Mechanism of Pattern Recognition Unaffected by Shift in Position — in: Biological Cybernetics 36 (1980) pp193–202Google Scholar
- [Heh87]HEHENBERGER,G.: Ähnlichkeitserkennung und Fehlerkorrektur im Assoziativspeichersystem NEUNET-3, Kepler University Linz (1987)Google Scholar
- [Hop82]HOPFIELD,J.J.: Neural networks and physical systems with emergent collective computational abilities — in: Proc. of the National Academy of Sciences, USA, 79 (1982)Google Scholar
- [Rei78]REICHL,E.R.: Neuronennetz-ähnliche Strukturen als Assoziativspeicher — in: Applied Computer Science, Vol 8: Datenstrukturen, Graphen, Algorithmen (1978)Google Scholar
- [Rum86]RUMELHART, D.E., HINTON, G.E. & WILLIAMS, R.J.: Learning Internal Representations by Error Propagation — in: Rumelhart, D.E. & McClelland, J.L. (eds.) Parallel Distributed Processing Vol I, MIT Press Cambridge, MA. (1986)Google Scholar