Title :
MAN: mass attraction network
Author :
Erdem, Mahmut Hilmi ; Ozturk, Yusuf
Author_Institution :
Dept. of Comput. Eng., Ege Univ., Izmir, Turkey
fDate :
30 May-2 Jun 1994
Abstract :
In this study, a binary associative memory, inspired from Newton´s mass attraction theory is proposed and some related analysis is given. In the model, memory items are considered as masses in the interior or at the corners of a hypercube. In recall, “attraction forces” are computed and the memory item, whose “force” is the greatest, becomes the output pattern. Since the operation of the model is highly parallel, the network is extremely fast. Retrieving a memory item takes only two steps. The proposed model has been observed to be superior to Hamming net, Hopfield network and Harmony theory in various aspects
Keywords :
content-addressable storage; learning (artificial intelligence); neural nets; parallel processing; MAN; attraction forces; binary associative memory; hypercube; mass attraction network; output pattern; parallel operation; Associative memory; Computational modeling; Computer simulation; Cost function; Equations; Gravity; Hamming distance; Information theory; Pattern analysis; Simulated annealing;
Conference_Titel :
Circuits and Systems, 1994. ISCAS '94., 1994 IEEE International Symposium on
Conference_Location :
London
Print_ISBN :
0-7803-1915-X
DOI :
10.1109/ISCAS.1994.409624
