1986 | OriginalPaper | Buchkapitel
A Physiological Neural Network as an Autoassociative Memory
verfasst von : J. Buhmann, K. Schulten
Erschienen in: Disordered Systems and Biological Organization
Verlag: Springer Berlin Heidelberg
Enthalten in: Professional Book Archive
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
We consider a neural network model in which the single neurons are chosen to resemble closely known physiological properties. The neurons are assumed to be linked by synapses which change their strength according to Hebbian rules [1] on a short time scale (100ms) [2]. Each nerve cell receives input from a primary set of receptors, which offer learning and test patterns without changing their own properties. The activity of the neurons is interpreted as the output of the network (see Fig.1). The backward bended arrows in Fig.1 indicate the feed-back due to the effect of the neuron activity on the synaptic strengths Sik between neuron k and i in the neural network.