Title :
Unstable vectors in Hopfield nets
Author :
Lai, W.K. ; Coghill, G.G.
Author_Institution :
Dept. of Electr. & Electron. Eng., Auckland Univ., New Zealand
Abstract :
It has been shown that the storage capacity of the Hopfield net is a function of the size of the network. However, it is also known that even if the number of input vectors is less than the maximum storage capacity of the network, incorrect associations can still occur. Such problems exist if these vectors have too large an overlap. The authors show some experimental results, as well as the mathematical analysis relating the stability of the input vectors to a similarity measure based on the Hamming distance
Keywords :
content-addressable storage; neural nets; Hamming distance; Hopfield nets; associative memories; error correction; incorrect associations; input vectors; mathematical analysis; random vectors; similarity measures; stability; storage capacity; unstable vectors; Artificial neural networks; Associative memory; Feeds; Hamming distance; Hopfield neural networks; Intelligent networks; Mathematical analysis; Neural networks; Neurons; Stability analysis;
Conference_Titel :
Circuits and Systems, 1991., IEEE International Sympoisum on
Print_ISBN :
0-7803-0050-5
DOI :
10.1109/ISCAS.1991.176042
