Title :
Information capacity of associative memories
Author :
Kuh, Anthony ; Dickinson, Bradley W.
Author_Institution :
Dept. of Electr. Eng., Hawaii Univ., Honolulu, HI, USA
fDate :
1/1/1989 12:00:00 AM
Abstract :
Associative memory networks consisting of highly interconnected binary-valued cells have been used to model neural networks. Tight asymptotic bounds have been found for the information capacity of these networks. The authors derive the asymptotic information capacity of these networks using results from normal approximation theory and theorems about exchangeable random variables
Keywords :
content-addressable storage; information theory; neural nets; approximation theory; associative memories; binary associative memory network; exchangeable random variables; highly interconnected binary-valued cells; information capacity; neural networks; tight asymptotic bounds; Approximation methods; Associative memory; Biological neural networks; Brain modeling; Hamming distance; Probes; Random variables; Stability;
Journal_Title :
Information Theory, IEEE Transactions on
