Architecture and Implementation of an Associative Memory Using Sparse Clustered Networks - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2012

Architecture and Implementation of an Associative Memory Using Sparse Clustered Networks

Résumé

Associative memories are alternatives to indexed memories that when implemented in hardware can benefit many applications such as data mining. The classical neural network based methodology is impractical to implement since in order to increase the size of the memory, the number of information bits stored per memory bit (efficiency) approaches zero. In addition, the length of a message to be stored and retrieved needs to be the same size as the number of nodes in the network causing the total number of messages the network is capable of storing (diversity) to be limited. Recently, a novel algorithm based on sparse clustered neural networks has been proposed that achieves nearly optimal efficiency and large diversity. In this paper, a proof-of-concept hardware implementation of these networks is presented. The limitations and possible future research areas are discussed.
Fichier non déposé

Dates et versions

hal-01056534 , version 1 (20-08-2014)

Identifiants

Citer

Hooman Jarollahi, Naoya Onizawa, Vincent Gripon, Warren J. Gross. Architecture and Implementation of an Associative Memory Using Sparse Clustered Networks. ISCAS 2012 : IEEE International Symposium on Circuits and Systems, May 2012, Seoul, South Korea. pp.2901-2904, ⟨10.1109/ISCAS.2012.6271922⟩. ⟨hal-01056534⟩
243 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More