Improving Storage of Patterns in Recurrent Neural Networks: Clones Based Model and Architecture

Hugues Nono Wouafo 1 Cyrille Chavet 2 Philippe Coussy 2
1 Lab-STICC_UBS_CACS_MOCS
Lab-STICC - Laboratoire des sciences et techniques de l'information, de la communication et de la connaissance
Abstract : Artificial neural networks are used in various domains like computer science and computer engineering for tasks like image processing, design of associative memories... The goal is to mimic the impressive brain ability to process or to memorize and retrieve information. Recently a new model of neural network has been proposed and has been applied to design associative memories. Even if this model seems to be really efficient, it suffers from many weaknesses. Some propositions have been made to address these problems but they are limited. In this paper, we propose a new concept in the field of binary neural networks to efficiently solve these problems while optimizing the cost of the architecture.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01101580
Contributor : Cyrille Chavet <>
Submitted on : Friday, January 9, 2015 - 9:46:38 AM
Last modification on : Monday, February 25, 2019 - 3:14:11 PM

Identifiers

  • HAL Id : hal-01101580, version 1

Citation

Hugues Nono Wouafo, Cyrille Chavet, Philippe Coussy. Improving Storage of Patterns in Recurrent Neural Networks: Clones Based Model and Architecture. IEEE Int'l Symposium on Circuits & Systems (ISCAS), May 2015, Lisbonne, Portugal. ⟨hal-01101580⟩

Share

Metrics

Record views

1635