Clone-Based Encoded Neural Networks to Design Efficient Associative Memories

Hugues Wouafo (1) , Cyrille Chavet (2, 3) , Philippe Coussy (3)

1 Lab-STICC_UBS_CACS_MOCS
2 Lab-STICC - Laboratoire des sciences et techniques de l'information, de la communication et de la connaissance
3 Lab-STICC_UBS_CACS_MOCS

In this paper, we introduce a neural network model named Clone based Neural Network (CbNN) to design associative memories. Neurons in CbNN can be cloned statically or dynamically which allows to increase the number of data that can be stored and retrieved. Thanks to their plasticity, CbNN can handle correlated information more robustly than existing models and thus provides better memory capacity. We experiment this model in Encoded Neural Networks also known as Gripon-Berrou neural networks. Numerical simulations demonstrate that memory and recall abilities of CbNN outperform state of art for the same memory footprint.

Mots clés: Index Terms-Associative Memories; Neural Networks; Content Addressable Memories

Domaines: Architectures Matérielles [cs.AR]; Systèmes embarqués; Traitement du signal et de l'image [eess.SP]; Réseau de neurones [cs.NE]; Ingénierie assistée par ordinateur

Hugues Wouafo, Cyrille Chavet, Philippe Coussy. Clone-Based Encoded Neural Networks to Design Efficient Associative Memories. IEEE Transactions on Neural Networks and Learning Systems, 2019, 30 (10), pp.1-14. ⟨10.1109/TNNLS.2018.2890658⟩. ⟨hal-02060021⟩