Restricted Boltzmann machine associative memory

Koki Nagatani, Masafumi Hagiwara

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Restricted Boltzmann machine associative memory (RBMAM) is proposed in this paper. RBMAM memorizes patterns using contrastive divergence learning procedure. It recalls by calculating the reconstruction of pattern using conditional probability. In order to examine the performance of the proposed RBMAM, extensive computer simulations have been carried out. As the result, it has shown that the performance of RBMAM is overwhelming compared with the conventional neural network associative memories. For example as for storage capacity, RBMAM can store about from 2N<inf>hidden</inf> to AN<inf>hideen</inf> patterns, where N<inf>hidden</inf> denotes the number of neurons in the hidden layer. Similarly we have obtained superior performance of RBMAM in respect of noise tolerance and pattern complement.

Original languageEnglish
Title of host publicationProceedings of the International Joint Conference on Neural Networks
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages3745-3750
Number of pages6
ISBN (Print)9781479914845
DOIs
Publication statusPublished - 2014 Sep 3
Event2014 International Joint Conference on Neural Networks, IJCNN 2014 - Beijing, China
Duration: 2014 Jul 62014 Jul 11

Other

Other2014 International Joint Conference on Neural Networks, IJCNN 2014
CountryChina
CityBeijing
Period14/7/614/7/11

Fingerprint

Data storage equipment
Neurons
Neural networks
Computer simulation

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Cite this

Nagatani, K., & Hagiwara, M. (2014). Restricted Boltzmann machine associative memory. In Proceedings of the International Joint Conference on Neural Networks (pp. 3745-3750). [6889573] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IJCNN.2014.6889573

Restricted Boltzmann machine associative memory. / Nagatani, Koki; Hagiwara, Masafumi.

Proceedings of the International Joint Conference on Neural Networks. Institute of Electrical and Electronics Engineers Inc., 2014. p. 3745-3750 6889573.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nagatani, K & Hagiwara, M 2014, Restricted Boltzmann machine associative memory. in Proceedings of the International Joint Conference on Neural Networks., 6889573, Institute of Electrical and Electronics Engineers Inc., pp. 3745-3750, 2014 International Joint Conference on Neural Networks, IJCNN 2014, Beijing, China, 14/7/6. https://doi.org/10.1109/IJCNN.2014.6889573
Nagatani K, Hagiwara M. Restricted Boltzmann machine associative memory. In Proceedings of the International Joint Conference on Neural Networks. Institute of Electrical and Electronics Engineers Inc. 2014. p. 3745-3750. 6889573 https://doi.org/10.1109/IJCNN.2014.6889573
Nagatani, Koki ; Hagiwara, Masafumi. / Restricted Boltzmann machine associative memory. Proceedings of the International Joint Conference on Neural Networks. Institute of Electrical and Electronics Engineers Inc., 2014. pp. 3745-3750
@inproceedings{30f27bb56e8347fd9da90386b0fba5a6,
title = "Restricted Boltzmann machine associative memory",
abstract = "Restricted Boltzmann machine associative memory (RBMAM) is proposed in this paper. RBMAM memorizes patterns using contrastive divergence learning procedure. It recalls by calculating the reconstruction of pattern using conditional probability. In order to examine the performance of the proposed RBMAM, extensive computer simulations have been carried out. As the result, it has shown that the performance of RBMAM is overwhelming compared with the conventional neural network associative memories. For example as for storage capacity, RBMAM can store about from 2Nhidden to ANhideen patterns, where Nhidden denotes the number of neurons in the hidden layer. Similarly we have obtained superior performance of RBMAM in respect of noise tolerance and pattern complement.",
author = "Koki Nagatani and Masafumi Hagiwara",
year = "2014",
month = "9",
day = "3",
doi = "10.1109/IJCNN.2014.6889573",
language = "English",
isbn = "9781479914845",
pages = "3745--3750",
booktitle = "Proceedings of the International Joint Conference on Neural Networks",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Restricted Boltzmann machine associative memory

AU - Nagatani, Koki

AU - Hagiwara, Masafumi

PY - 2014/9/3

Y1 - 2014/9/3

N2 - Restricted Boltzmann machine associative memory (RBMAM) is proposed in this paper. RBMAM memorizes patterns using contrastive divergence learning procedure. It recalls by calculating the reconstruction of pattern using conditional probability. In order to examine the performance of the proposed RBMAM, extensive computer simulations have been carried out. As the result, it has shown that the performance of RBMAM is overwhelming compared with the conventional neural network associative memories. For example as for storage capacity, RBMAM can store about from 2Nhidden to ANhideen patterns, where Nhidden denotes the number of neurons in the hidden layer. Similarly we have obtained superior performance of RBMAM in respect of noise tolerance and pattern complement.

AB - Restricted Boltzmann machine associative memory (RBMAM) is proposed in this paper. RBMAM memorizes patterns using contrastive divergence learning procedure. It recalls by calculating the reconstruction of pattern using conditional probability. In order to examine the performance of the proposed RBMAM, extensive computer simulations have been carried out. As the result, it has shown that the performance of RBMAM is overwhelming compared with the conventional neural network associative memories. For example as for storage capacity, RBMAM can store about from 2Nhidden to ANhideen patterns, where Nhidden denotes the number of neurons in the hidden layer. Similarly we have obtained superior performance of RBMAM in respect of noise tolerance and pattern complement.

UR - http://www.scopus.com/inward/record.url?scp=84908471909&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84908471909&partnerID=8YFLogxK

U2 - 10.1109/IJCNN.2014.6889573

DO - 10.1109/IJCNN.2014.6889573

M3 - Conference contribution

AN - SCOPUS:84908471909

SN - 9781479914845

SP - 3745

EP - 3750

BT - Proceedings of the International Joint Conference on Neural Networks

PB - Institute of Electrical and Electronics Engineers Inc.

ER -