Title :
On the reduction of connections in Hopfield associative memories
Author :
Tang, Zhimin ; Li, Guo-Jie
Author_Institution :
Inst. of Comput Technol., Acad. Sinica, Beijing, China
Abstract :
The attempt to reduce the number of interconnections in artificial neural networks has been taken by many researchers to ease the real implementation. Since such reduction usually degrades the performance of ANS, a trade-off between the ease of implementation and the network performance must be made. The authors prove, for a Hopfield neural network, that such a tradeoff is impossible, i.e., the reduction is fruitless
Keywords :
content-addressable storage; neural nets; Hopfield associative memories; artificial neural networks; connections reduction; interconnections; network performance; Artificial neural networks; Associative memory; Computers; Degradation; Hamming distance; Hopfield neural networks; Intelligent networks; Neural networks; Neurons; Probes;
Conference_Titel :
Circuits and Systems, 1991. Conference Proceedings, China., 1991 International Conference on
Conference_Location :
Shenzhen
DOI :
10.1109/CICCAS.1991.184340
