Title :
Tuning capability of binary perceptron
Author :
Biswas, N.N. ; Bhattacharyya, S.K.
Author_Institution :
Dept. of Electr. & Commun. Eng., Indian Inst. of Sci., Bangalore, India
fDate :
6/20/1991 12:00:00 AM
Abstract :
A theorem called the tuning theorem, which brings out the tuning capability of perceptrons where both the inputs and outputs are binary, has been stated and proved. It has been shown how with the help of this property, an associative memory has been developed with a single layer of perceptrons working as a massively parallel machine, rather than as a fully connected iterative Hopfield network.
Keywords :
content-addressable storage; memory architecture; neural nets; parallel architectures; tuning; CAM; associative memory; binary perceptron; massively parallel machine; neural networks; tuning capability; tuning theorem;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19910751