Title :
Computation of the distributed associative memory and numerical stability of the generalized inverse matrix
Author :
Mohideen, S. ; Cherkassky, Vladimir ; Wechsler, Harry
Author_Institution :
Hewlett-Packard, Cupertino, CA, USA
Abstract :
Summary form only given. A novel recursive algorithm to calculate the recursive generalized inverse (RGI) needed to learn some distributed associative memory is introduced. The authors show that this new RGI algorithm provides better numerical stability when compared with Greville´s algorithm. The RGI has been applied successfully to fault-tolerant information retrieval tasks.<>
Keywords :
content-addressable storage; convergence of numerical methods; distributed processing; matrix algebra; Greville´s algorithm; distributed associative memory; fault-tolerant information retrieval; generalized inverse matrix; numerical stability; recursive generalized inverse; Associative memories; Convergence of numerical methods; Distributed computing; Matrices;
Conference_Titel :
Neural Networks, 1989. IJCNN., International Joint Conference on
Conference_Location :
Washington, DC, USA
DOI :
10.1109/IJCNN.1989.118340
