A full-parallel implementation of Self-Organizing Maps on hardware

Leonardo A. Dias; Augusto M. P. Damasceno; Elena Gaura; Marcelo A.C. Fernandes

doi:10.1016/j.neunet.2021.05.021

A full-parallel implementation of Self-Organizing Maps on hardware

Leonardo A. Dias, Augusto M. P. Damasceno, Elena Gaura, Marcelo A.C. Fernandes

Complex Systems

Research output: Contribution to journal › Article › peer-review

7 Citations (Scopus)

115 Downloads (Pure)

Abstract

Self-Organizing Maps (SOMs) are extensively used for data clustering and dimensionality reduction. However, if applications are to fully benefit from SOM based techniques, high-speed processing is demanding, given that data tends to be both highly dimensional and yet “big”. Hence, a fully parallel architecture for the SOM is introduced to optimize the system’s data processing time. Unlike most literature approaches, the architecture proposed here does not contain sequential steps - a common limiting factor for processing speed. The architecture was validated on FPGA and evaluated concerning hardware throughput and the use of resources. Comparisons to the state of the art show a speedup of 8.91x over a partially serial implementation, using less than 15% of hardware resources available. Thus, the method proposed here points to a hardware architecture that will not be obsolete quickly.

Original language	English
Pages (from-to)	818-827
Number of pages	10
Journal	Neural Networks
Volume	143
Early online date	21 May 2021
DOIs	https://doi.org/10.1016/j.neunet.2021.05.021
Publication status	Published - Nov 2021

Bibliographical note

NOTICE: this is the author’s version of a work that was accepted for publication in Neural Networks. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in Neural Networks, 143, (2021) DOI: 10.1016/j.neunet.2021.05.021

© 2021, Elsevier. Licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International http://creativecommons.org/licenses/by-nc-nd/4.0/

Funder

Funded in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) - Institutional Program for Internationalization (CAPES - PrInt), Brazil and EPSRC Grant EP/S031863/1.

Keywords

FPGA
Hardware
Parallel design
Self-Organizing Map

ASJC Scopus subject areas

Cognitive Neuroscience
Artificial Intelligence

Access to Document

10.1016/j.neunet.2021.05.021Licence: CC BY

Dias_et_al_Full-parallel_implementationFinal published version, 1.34 MBLicence: CC BY

Cite this

@article{cd069a7230a645f38434d197bc2bd862,

title = "A full-parallel implementation of Self-Organizing Maps on hardware",

abstract = "Self-Organizing Maps (SOMs) are extensively used for data clustering and dimensionality reduction. However, if applications are to fully benefit from SOM based techniques, high-speed processing is demanding, given that data tends to be both highly dimensional and yet “big”. Hence, a fully parallel architecture for the SOM is introduced to optimize the system{\textquoteright}s data processing time. Unlike most literature approaches, the architecture proposed here does not contain sequential steps - a common limiting factor for processing speed. The architecture was validated on FPGA and evaluated concerning hardware throughput and the use of resources. Comparisons to the state of the art show a speedup of 8.91x over a partially serial implementation, using less than 15% of hardware resources available. Thus, the method proposed here points to a hardware architecture that will not be obsolete quickly.",

keywords = "FPGA, Hardware, Parallel design, Self-Organizing Map",

author = "Dias, {Leonardo A.} and Damasceno, {Augusto M. P.} and Elena Gaura and Fernandes, {Marcelo A.C.}",

note = "NOTICE: this is the author{\textquoteright}s version of a work that was accepted for publication in Neural Networks. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in Neural Networks, 143, (2021) DOI: 10.1016/j.neunet.2021.05.021 {\textcopyright} 2021, Elsevier. Licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International http://creativecommons.org/licenses/by-nc-nd/4.0/ ",

year = "2021",

month = nov,

doi = "10.1016/j.neunet.2021.05.021",

language = "English",

volume = "143",

pages = "818--827",

journal = "Neural Networks",

issn = "0893-6080",

publisher = "Elsevier",

}

TY - JOUR

T1 - A full-parallel implementation of Self-Organizing Maps on hardware

AU - Dias, Leonardo A.

AU - Damasceno, Augusto M. P.

AU - Gaura, Elena

AU - Fernandes, Marcelo A.C.

N1 - NOTICE: this is the author’s version of a work that was accepted for publication in Neural Networks. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in Neural Networks, 143, (2021) DOI: 10.1016/j.neunet.2021.05.021 © 2021, Elsevier. Licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International http://creativecommons.org/licenses/by-nc-nd/4.0/

PY - 2021/11

Y1 - 2021/11

N2 - Self-Organizing Maps (SOMs) are extensively used for data clustering and dimensionality reduction. However, if applications are to fully benefit from SOM based techniques, high-speed processing is demanding, given that data tends to be both highly dimensional and yet “big”. Hence, a fully parallel architecture for the SOM is introduced to optimize the system’s data processing time. Unlike most literature approaches, the architecture proposed here does not contain sequential steps - a common limiting factor for processing speed. The architecture was validated on FPGA and evaluated concerning hardware throughput and the use of resources. Comparisons to the state of the art show a speedup of 8.91x over a partially serial implementation, using less than 15% of hardware resources available. Thus, the method proposed here points to a hardware architecture that will not be obsolete quickly.

AB - Self-Organizing Maps (SOMs) are extensively used for data clustering and dimensionality reduction. However, if applications are to fully benefit from SOM based techniques, high-speed processing is demanding, given that data tends to be both highly dimensional and yet “big”. Hence, a fully parallel architecture for the SOM is introduced to optimize the system’s data processing time. Unlike most literature approaches, the architecture proposed here does not contain sequential steps - a common limiting factor for processing speed. The architecture was validated on FPGA and evaluated concerning hardware throughput and the use of resources. Comparisons to the state of the art show a speedup of 8.91x over a partially serial implementation, using less than 15% of hardware resources available. Thus, the method proposed here points to a hardware architecture that will not be obsolete quickly.

KW - FPGA

KW - Hardware

KW - Parallel design

KW - Self-Organizing Map

UR - http://www.scopus.com/inward/record.url?scp=85107586103&partnerID=8YFLogxK

U2 - 10.1016/j.neunet.2021.05.021

DO - 10.1016/j.neunet.2021.05.021

M3 - Article

SN - 0893-6080

VL - 143

SP - 818

EP - 827

JO - Neural Networks

JF - Neural Networks

ER -

A full-parallel implementation of Self-Organizing Maps on hardware

Abstract

Bibliographical note

Funder

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this