© The Institution of Electrical Engineers
The Hamming network is a binary associative memory which exhibits the maximum theoretically obtainable performance in terms of capacity and associativity. The structure is not ideal and two modifications which reduce the number of connections in the correlation matrix and in the selection layers are suggested. These modifications offer a significant reduction in the number of calculations required.
References
-
-
1)
-
R.P. Lippman
.
An introduction to computing with neural nets.
IEEE ASSP Mag.
,
4 -
22
-
2)
-
P. Wasserman
.
(1989)
, Neural computing theory and practice.
-
3)
-
P.K. Houselander ,
J.T. Taylor
.
Comparing performance of Hebbian and delta trained Hopfield networks.
Electron. Lett.
,
2 ,
85 -
87
-
4)
-
Hopfield, J.J.: `Neural networks and physical systems with emergent collective computational abilities', Proc. Natl. Acad. Sci., 1982, 79, p. 2554–2558.
http://iet.metastore.ingenta.com/content/journals/10.1049/el_19900460
Related content
content/journals/10.1049/el_19900460
pub_keyword,iet_inspecKeyword,pub_concept
6
6