Title :
Optimal noise rejection in linear associative memories
Author :
Olivier, Philip D.
Author_Institution :
Div. of Eng., Texas Univ., San Antonio, TX, USA
Abstract :
An associative memory, designed to recall a data vector when presented with the appropriate uncorrupted key vector and to optimally extract the data vector when presented with a corrupted key vector, is described. The extraction is optimal in the sense that a quadratic cost functional is minimized
Keywords :
content-addressable storage; optimisation; data vector; key vector; linear associative memories; optimal noise rejection; quadratic cost functional; Associative memory; Cost function; Data mining; Statistics; Terminology; Vectors;
Journal_Title :
Systems, Man and Cybernetics, IEEE Transactions on
