A scalable multi-signal approach for the parallelization of self-organizing neural networks
Self-Organizing Neural Networks (SONNs) have a wide range of applications with massive computational requirements that often need to be satisfied with optimized parallel algorithms and implementations. In literature, SONN have been generally parallelized with GPU computing according to a single-sign...
Uloženo v:
| Vydáno v: | Neural networks Ročník 123; s. 108 - 117 |
|---|---|
| Hlavní autoři: | , , , |
| Médium: | Journal Article |
| Jazyk: | angličtina |
| Vydáno: |
United States
Elsevier Ltd
01.03.2020
|
| Témata: | |
| ISSN: | 0893-6080, 1879-2782, 1879-2782 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Abstract | Self-Organizing Neural Networks (SONNs) have a wide range of applications with massive computational requirements that often need to be satisfied with optimized parallel algorithms and implementations. In literature, SONN have been generally parallelized with GPU computing according to a single-signal paradigm: each GPU thread manages one or more nodes of the network and works concurrently on one input signal at the time. This paper presents two contributions. The first one is the experimental proof that the single-signal approach for SONNs is not optimal for the task, as it is intrinsically sequential at its core and thus inherently limited in its performance. The non-optimality of the single-signal paradigm is illustrated via a specific and simplified benchmark. The second contribution is the introduction of a new multi-signal paradigm for the parallelization of SONNs, whereby multiple signals are processed at once in each iteration hence allowing different GPU threads to work on different signals. The advantages of the multi-signal approach are shown through several benchmarks involving the Self-Organizing Adaptive Map (SOAM) algorithm as a basis for evaluation. Having a graph-based termination condition that depends on the features of the network being grown, the SOAM algorithm allows assessing both functional equivalence and performances of the paradigm proposed without relying on arbitrary thresholds. Nonetheless, the evaluation proposed has a broader scope since it refers to a unified framework for the GPU parallelization of a generic SONN.
•The topic is the parallelization of self-organizing neural networks (SONNs).•A single-signal approach is common for GPU implementations.•The paper shows that single-signal is inherently sequential and has limited scalability.•The paper thus proposes an effective, scalable, multi-signal approach for the parallelization of SONNs.•The multi-signal performance is validated with a carefully-constructed benchmark on SOAM. |
|---|---|
| AbstractList | Self-Organizing Neural Networks (SONNs) have a wide range of applications with massive computational requirements that often need to be satisfied with optimized parallel algorithms and implementations. In literature, SONN have been generally parallelized with GPU computing according to a single-signal paradigm: each GPU thread manages one or more nodes of the network and works concurrently on one input signal at the time. This paper presents two contributions. The first one is the experimental proof that the single-signal approach for SONNs is not optimal for the task, as it is intrinsically sequential at its core and thus inherently limited in its performance. The non-optimality of the single-signal paradigm is illustrated via a specific and simplified benchmark. The second contribution is the introduction of a new multi-signal paradigm for the parallelization of SONNs, whereby multiple signals are processed at once in each iteration hence allowing different GPU threads to work on different signals. The advantages of the multi-signal approach are shown through several benchmarks involving the Self-Organizing Adaptive Map (SOAM) algorithm as a basis for evaluation. Having a graph-based termination condition that depends on the features of the network being grown, the SOAM algorithm allows assessing both functional equivalence and performances of the paradigm proposed without relying on arbitrary thresholds. Nonetheless, the evaluation proposed has a broader scope since it refers to a unified framework for the GPU parallelization of a generic SONN. Self-Organizing Neural Networks (SONNs) have a wide range of applications with massive computational requirements that often need to be satisfied with optimized parallel algorithms and implementations. In literature, SONN have been generally parallelized with GPU computing according to a single-signal paradigm: each GPU thread manages one or more nodes of the network and works concurrently on one input signal at the time. This paper presents two contributions. The first one is the experimental proof that the single-signal approach for SONNs is not optimal for the task, as it is intrinsically sequential at its core and thus inherently limited in its performance. The non-optimality of the single-signal paradigm is illustrated via a specific and simplified benchmark. The second contribution is the introduction of a new multi-signal paradigm for the parallelization of SONNs, whereby multiple signals are processed at once in each iteration hence allowing different GPU threads to work on different signals. The advantages of the multi-signal approach are shown through several benchmarks involving the Self-Organizing Adaptive Map (SOAM) algorithm as a basis for evaluation. Having a graph-based termination condition that depends on the features of the network being grown, the SOAM algorithm allows assessing both functional equivalence and performances of the paradigm proposed without relying on arbitrary thresholds. Nonetheless, the evaluation proposed has a broader scope since it refers to a unified framework for the GPU parallelization of a generic SONN.Self-Organizing Neural Networks (SONNs) have a wide range of applications with massive computational requirements that often need to be satisfied with optimized parallel algorithms and implementations. In literature, SONN have been generally parallelized with GPU computing according to a single-signal paradigm: each GPU thread manages one or more nodes of the network and works concurrently on one input signal at the time. This paper presents two contributions. The first one is the experimental proof that the single-signal approach for SONNs is not optimal for the task, as it is intrinsically sequential at its core and thus inherently limited in its performance. The non-optimality of the single-signal paradigm is illustrated via a specific and simplified benchmark. The second contribution is the introduction of a new multi-signal paradigm for the parallelization of SONNs, whereby multiple signals are processed at once in each iteration hence allowing different GPU threads to work on different signals. The advantages of the multi-signal approach are shown through several benchmarks involving the Self-Organizing Adaptive Map (SOAM) algorithm as a basis for evaluation. Having a graph-based termination condition that depends on the features of the network being grown, the SOAM algorithm allows assessing both functional equivalence and performances of the paradigm proposed without relying on arbitrary thresholds. Nonetheless, the evaluation proposed has a broader scope since it refers to a unified framework for the GPU parallelization of a generic SONN. Self-Organizing Neural Networks (SONNs) have a wide range of applications with massive computational requirements that often need to be satisfied with optimized parallel algorithms and implementations. In literature, SONN have been generally parallelized with GPU computing according to a single-signal paradigm: each GPU thread manages one or more nodes of the network and works concurrently on one input signal at the time. This paper presents two contributions. The first one is the experimental proof that the single-signal approach for SONNs is not optimal for the task, as it is intrinsically sequential at its core and thus inherently limited in its performance. The non-optimality of the single-signal paradigm is illustrated via a specific and simplified benchmark. The second contribution is the introduction of a new multi-signal paradigm for the parallelization of SONNs, whereby multiple signals are processed at once in each iteration hence allowing different GPU threads to work on different signals. The advantages of the multi-signal approach are shown through several benchmarks involving the Self-Organizing Adaptive Map (SOAM) algorithm as a basis for evaluation. Having a graph-based termination condition that depends on the features of the network being grown, the SOAM algorithm allows assessing both functional equivalence and performances of the paradigm proposed without relying on arbitrary thresholds. Nonetheless, the evaluation proposed has a broader scope since it refers to a unified framework for the GPU parallelization of a generic SONN. •The topic is the parallelization of self-organizing neural networks (SONNs).•A single-signal approach is common for GPU implementations.•The paper shows that single-signal is inherently sequential and has limited scalability.•The paper thus proposes an effective, scalable, multi-signal approach for the parallelization of SONNs.•The multi-signal performance is validated with a carefully-constructed benchmark on SOAM. |
| Author | Cantoni, Virginio Musci, Mirto Piastra, Marco Parigi, Giacomo |
| Author_xml | – sequence: 1 givenname: Mirto orcidid: 0000-0002-3868-3738 surname: Musci fullname: Musci, Mirto email: mirto.musci@unipv.it – sequence: 2 givenname: Giacomo surname: Parigi fullname: Parigi, Giacomo email: giacomo.parigi@unipv.it – sequence: 3 givenname: Virginio surname: Cantoni fullname: Cantoni, Virginio email: virginio.cantoni@unipv.it – sequence: 4 givenname: Marco surname: Piastra fullname: Piastra, Marco email: marco.piastra@unipv.it |
| BackLink | https://www.ncbi.nlm.nih.gov/pubmed/31838325$$D View this record in MEDLINE/PubMed |
| BookMark | eNqFkD1vFDEQhi0URC6Bf4CQS5pd_MWuTYEURXxJkWigorBmvbMXHz77sL1B5NfjcElDAdVIo-d9NfOckZOYIhLynLOeMz682vUR14i1F4ybnvO-LR-RDdej6cSoxQnZMG1kNzDNTslZKTvG2KCVfEJOJddSS_F6Q75d0OIgwBSQ7tdQfVf8NkKgcDjkBO6aLinTeo30ABlCwOBvofoUaVpowbB0KW8h-lsft7Qd1Jg26s-Uv5en5PECoeCz-3lOvr5_9-XyY3f1-cOny4urzslB1A6kNsOsNANYRqfGUc6D5jNn0kyT0hMgd2j4otSitDNczhObmQYxOD0wZuQ5eXnsbRf_WLFUu_fFYQgQMa3FCilGqbkwqqEv7tF12uNsD9nvIf-yD0Ia8OYIuJxKybhY5-ufh2sGHyxn9s6-3dmjfXtn33Ju27KF1V_hh_7_xN4eY9gk3XjMtjiP0eHsM7pq5-T_XfAbQXOhdw |
| CitedBy_id | crossref_primary_10_1016_j_neunet_2021_05_021 crossref_primary_10_1155_2021_4105784 crossref_primary_10_1155_2022_5141745 crossref_primary_10_3390_s22176449 |
| Cites_doi | 10.1007/s10766-015-0395-0 10.6109/jicce.2014.12.2.128 10.1016/j.asoc.2012.01.018 10.1007/s12541-012-0179-z 10.1016/j.patrec.2009.09.009 10.1016/S0893-6080(02)00078-3 10.3233/ICA-2010-0334 10.1016/0893-6080(94)90091-4 10.1016/j.jpdc.2012.05.008 10.1016/j.neucom.2018.04.015 10.1016/S0893-6080(05)80024-3 10.1109/SURV.2011.040310.00002 |
| ContentType | Journal Article |
| Copyright | 2019 Elsevier Ltd Copyright © 2019 Elsevier Ltd. All rights reserved. |
| Copyright_xml | – notice: 2019 Elsevier Ltd – notice: Copyright © 2019 Elsevier Ltd. All rights reserved. |
| DBID | AAYXX CITATION NPM 7X8 |
| DOI | 10.1016/j.neunet.2019.11.016 |
| DatabaseName | CrossRef PubMed MEDLINE - Academic |
| DatabaseTitle | CrossRef PubMed MEDLINE - Academic |
| DatabaseTitleList | PubMed MEDLINE - Academic |
| Database_xml | – sequence: 1 dbid: NPM name: PubMed url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed sourceTypes: Index Database – sequence: 2 dbid: 7X8 name: MEDLINE - Academic url: https://search.proquest.com/medline sourceTypes: Aggregation Database |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Computer Science |
| EISSN | 1879-2782 |
| EndPage | 117 |
| ExternalDocumentID | 31838325 10_1016_j_neunet_2019_11_016 S0893608019303752 |
| Genre | Journal Article |
| GroupedDBID | --- --K --M -~X .DC .~1 0R~ 123 186 1B1 1RT 1~. 1~5 29N 4.4 457 4G. 53G 5RE 5VS 6TJ 7-5 71M 8P~ 9JM 9JN AABNK AACTN AADPK AAEDT AAEDW AAIAV AAIKJ AAKOC AALRI AAOAW AAQFI AAQXK AAXLA AAXUO AAYFN ABAOU ABBOA ABCQJ ABEFU ABFNM ABFRF ABHFT ABIVO ABJNI ABLJU ABMAC ABXDB ABYKQ ACAZW ACDAQ ACGFO ACGFS ACIUM ACNNM ACRLP ACZNC ADBBV ADEZE ADGUI ADJOM ADMUD ADRHT AEBSH AECPX AEFWE AEKER AENEX AFKWA AFTJW AFXIZ AGHFR AGUBO AGWIK AGYEJ AHHHB AHJVU AHZHX AIALX AIEXJ AIKHN AITUG AJBFU AJOXV ALMA_UNASSIGNED_HOLDINGS AMFUW AMRAJ AOUOD ARUGR ASPBG AVWKF AXJTR AZFZN BJAXD BKOJK BLXMC CS3 DU5 EBS EFJIC EFLBG EJD EO8 EO9 EP2 EP3 F0J F5P FDB FEDTE FGOYB FIRID FNPLU FYGXN G-2 G-Q G8K GBLVA GBOLZ HLZ HMQ HVGLF HZ~ IHE J1W JJJVA K-O KOM KZ1 LG9 LMP M2V M41 MHUIS MO0 MOBAO MVM N9A O-L O9- OAUVE OZT P-8 P-9 P2P PC. Q38 R2- RIG ROL RPZ SBC SCC SDF SDG SDP SES SEW SNS SPC SPCBC SSN SST SSV SSW SSZ T5K TAE UAP UNMZH VOH WUQ XPP ZMT ~G- 9DU AATTM AAXKI AAYWO AAYXX ABDPE ABWVN ACLOT ACRPL ACVFH ADCNI ADNMO AEIPS AEUPX AFJKZ AFPUW AGQPQ AIGII AIIUN AKBMS AKRWK AKYEP ANKPU APXCP CITATION EFKBS ~HD NPM PKN 7X8 |
| ID | FETCH-LOGICAL-c362t-a3896d480aaf7c4773d681d1039bb48bae1ce91f44f48c913db0d08a26c860093 |
| ISICitedReferencesCount | 4 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000511985000009&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0893-6080 1879-2782 |
| IngestDate | Sun Sep 28 03:18:31 EDT 2025 Wed Feb 19 02:31:53 EST 2025 Tue Nov 18 22:00:52 EST 2025 Sat Nov 29 07:19:45 EST 2025 Fri Feb 23 02:49:25 EST 2024 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Keywords | Surface reconstruction GPU parallelization Self-organizing adaptive map Self-organizing neural networks |
| Language | English |
| License | Copyright © 2019 Elsevier Ltd. All rights reserved. |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-c362t-a3896d480aaf7c4773d681d1039bb48bae1ce91f44f48c913db0d08a26c860093 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
| ORCID | 0000-0002-3868-3738 |
| PMID | 31838325 |
| PQID | 2327381294 |
| PQPubID | 23479 |
| PageCount | 10 |
| ParticipantIDs | proquest_miscellaneous_2327381294 pubmed_primary_31838325 crossref_citationtrail_10_1016_j_neunet_2019_11_016 crossref_primary_10_1016_j_neunet_2019_11_016 elsevier_sciencedirect_doi_10_1016_j_neunet_2019_11_016 |
| PublicationCentury | 2000 |
| PublicationDate | March 2020 2020-03-00 2020-Mar 20200301 |
| PublicationDateYYYYMMDD | 2020-03-01 |
| PublicationDate_xml | – month: 03 year: 2020 text: March 2020 |
| PublicationDecade | 2020 |
| PublicationPlace | United States |
| PublicationPlace_xml | – name: United States |
| PublicationTitle | Neural networks |
| PublicationTitleAlternate | Neural Netw |
| PublicationYear | 2020 |
| Publisher | Elsevier Ltd |
| Publisher_xml | – name: Elsevier Ltd |
| References | Kanimozhi, Bindu (b14) 2013; 2 Vojáček, Dráždilová, Dvorskỳ (b35) 2017 Fritzke (b8) 1995 Luo, Liu, Wu (b20) 2005; vol. 1 Seiffert, Michaelis (b32) 2001 Croom, Rucker, Romano, Webster (b5) 2010 (pp. 652–660). Vojáček, Dvorskỳ (b36) 2013 Banković, Moya, Araujo, Fraga, Vallejo, de Goyeneche (b1) 2010; 17 Liu, Sun, Yao, Wang, Zheng, Liu (b19) 2018 Patra, Chattopadhyay, Bhaumik, Roy (b27) 2011 Jang, Park, Jung (b13) 2008 Piastra (b29) 2009 Fliege, Benn (b6) 2016 Campbell, Berglund, Streit (b3) 2005; 2005 Kohonen (b15) 2001 Fujii, Azumi, Nishio, Kato, Edahiro (b10) 2013 Zhongwen, Hongzhi, Zhengping, Xincai (b38) 2005 Li, Hu, Li, Wu, Yang (b18) 2016; 44 Caridakis, Karpouzis, Drosopoulos, Kollias (b4) 2010; 31 Kuo, Wang, Chen (b17) 2012; 12 Kulkarni, Forster, Venayagamoorthy (b16) 2011; 13 Sharma, Kim, Singh (b33) 2012; 13 Piastra (b30) 2012 Fritzke (b9) 1996 Parigi, Stramieri, Pau, Piastra (b26) 2014 Qi, C. R., Su, H., Mo, K., & Guibas, L. J. (2017). Pointnet: Deep learning on point sets for 3d classification and segmentation. In Gajdoš, Platoš (b11) 2013 Mici, Parisi, Wermter (b24) 2018; 307 Paul, Gupta (b28) 2013; 3 Fritzke (b7) 1994; 7 White (b37) 1992; 5 Bauer, Cook, Khailany (b2) 2011 García-Rodríguez, Angelopoulou, Morell, Orts, Psarrou, García-Chamizo (b12) 2011 Martinetz, Schulten (b22) 1991; 1 Orts, Garcia-Rodriguez, Viejo, Cazorla, Morell (b25) 2012 Marsland, Shapiro, Nehmzow (b21) 2002; 15 Sharma, Moon (b34) 2014; 12 McConnell, Sturgeon, Henry, Mayne, Hurley (b23) 2012; vol. 341 Martinetz (10.1016/j.neunet.2019.11.016_b22) 1991; 1 Parigi (10.1016/j.neunet.2019.11.016_b26) 2014 Bauer (10.1016/j.neunet.2019.11.016_b2) 2011 Jang (10.1016/j.neunet.2019.11.016_b13) 2008 Li (10.1016/j.neunet.2019.11.016_b18) 2016; 44 Luo (10.1016/j.neunet.2019.11.016_b20) 2005; vol. 1 Fritzke (10.1016/j.neunet.2019.11.016_b9) 1996 Piastra (10.1016/j.neunet.2019.11.016_b30) 2012 Vojáček (10.1016/j.neunet.2019.11.016_b36) 2013 Fritzke (10.1016/j.neunet.2019.11.016_b8) 1995 Kohonen (10.1016/j.neunet.2019.11.016_b15) 2001 10.1016/j.neunet.2019.11.016_b31 Seiffert (10.1016/j.neunet.2019.11.016_b32) 2001 Kulkarni (10.1016/j.neunet.2019.11.016_b16) 2011; 13 Croom (10.1016/j.neunet.2019.11.016_b5) 2010 Orts (10.1016/j.neunet.2019.11.016_b25) 2012 Fliege (10.1016/j.neunet.2019.11.016_b6) 2016 Caridakis (10.1016/j.neunet.2019.11.016_b4) 2010; 31 Piastra (10.1016/j.neunet.2019.11.016_b29) 2009 White (10.1016/j.neunet.2019.11.016_b37) 1992; 5 Sharma (10.1016/j.neunet.2019.11.016_b34) 2014; 12 Kuo (10.1016/j.neunet.2019.11.016_b17) 2012; 12 Marsland (10.1016/j.neunet.2019.11.016_b21) 2002; 15 Campbell (10.1016/j.neunet.2019.11.016_b3) 2005; 2005 Gajdoš (10.1016/j.neunet.2019.11.016_b11) 2013 Paul (10.1016/j.neunet.2019.11.016_b28) 2013; 3 Sharma (10.1016/j.neunet.2019.11.016_b33) 2012; 13 Liu (10.1016/j.neunet.2019.11.016_b19) 2018 Mici (10.1016/j.neunet.2019.11.016_b24) 2018; 307 Zhongwen (10.1016/j.neunet.2019.11.016_b38) 2005 Banković (10.1016/j.neunet.2019.11.016_b1) 2010; 17 Fritzke (10.1016/j.neunet.2019.11.016_b7) 1994; 7 Patra (10.1016/j.neunet.2019.11.016_b27) 2011 Fujii (10.1016/j.neunet.2019.11.016_b10) 2013 García-Rodríguez (10.1016/j.neunet.2019.11.016_b12) 2011 Kanimozhi (10.1016/j.neunet.2019.11.016_b14) 2013; 2 McConnell (10.1016/j.neunet.2019.11.016_b23) 2012; vol. 341 Vojáček (10.1016/j.neunet.2019.11.016_b35) 2017 |
| References_xml | – reference: (pp. 652–660). – start-page: 275 year: 2013 end-page: 282 ident: b10 article-title: Data transfer matters for GPU computing publication-title: Parallel and distributed systems (ICPADS), 2013 international conference on – start-page: 408 year: 2013 end-page: 419 ident: b36 article-title: Growing neural gas–a parallel approach publication-title: Computer information systems and industrial management – volume: 1 start-page: 397 year: 1991 end-page: 402 ident: b22 article-title: A neural-gas network learns topologies publication-title: Artificial Neural Networks – volume: 12 start-page: 128 year: 2014 end-page: 134 ident: b34 article-title: GPU-Based optimization of self-organizing map feature matching for real-time stereo vision publication-title: Journal of Information and Communication Convergence Engineering – volume: 7 start-page: 1441 year: 1994 end-page: 1460 ident: b7 article-title: Growing cell structures–a self-organizing network for unsupervised and supervised learning publication-title: Neural Networks – volume: vol. 1 start-page: 622 year: 2005 end-page: 626 ident: b20 article-title: Artificial neural network computation on graphic process unit publication-title: Proceedings. 2005 IEEE international joint conference on – volume: 15 start-page: 1041 year: 2002 end-page: 1058 ident: b21 article-title: A self-organising network that grows when required publication-title: Neural Networks – start-page: 63 year: 2017 end-page: 72 ident: b35 article-title: Optimalization of parallel GNG by neurons assigned to processes publication-title: IFIP international conference on computer information systems and industrial management – volume: 2005 start-page: 5 year: 2005 end-page: 14 ident: b3 article-title: Graphics hardware implementation of the parameter-less self-organising map publication-title: Intelligent Data Engineering and Automated Learning-IDEAL – start-page: 231 year: 2013 end-page: 242 ident: b11 article-title: GPU Based parallelism for self-organizing map publication-title: Proceedings of the third international conference on intelligent human computer interaction (IHCI 2011), prague, czech republic, august, 2011 – volume: 5 start-page: 261 year: 1992 end-page: 275 ident: b37 article-title: Competitive hebbian learning: Algorithm and demonstrations publication-title: Neural Networks – start-page: 160 year: 2001 end-page: 166 ident: b32 article-title: Multi-dimensional self-organizing maps on massively parallel hardware publication-title: Advances in self-organising maps – volume: 12 year: 2012 ident: b17 article-title: Integration of growing self-organizing map and continuous genetic algorithm for grading lithium-ion battery cells publication-title: Applied Soft Computing – start-page: 4591 year: 2010 end-page: 4596 ident: b5 article-title: Visual sensing of continuum robot shape using self-organizing maps publication-title: Robotics and automation (ICRA), 2010 IEEE international conference on – year: 2012 ident: b25 article-title: GPGPU implementation of growing neural gas: Application to 3D scene reconstruction publication-title: Journal of Parallel and Distributed Computing – volume: 3 start-page: 288 year: 2013 end-page: 291 ident: b28 article-title: Image segmentation by self organizing map with mahalanobis distance publication-title: International Journal of Emerging Technology and Advanced Engineering – start-page: 1 year: 2011 end-page: 6 ident: b27 article-title: Using self organizing map in wireless sensor network for designing energy efficient topologies publication-title: Wireless communication, vehicular technology, information theory and aerospace & electronic systems technology (Wireless VITAE), 2011 2nd international conference on – volume: 17 start-page: 87 year: 2010 end-page: 102 ident: b1 article-title: Distributed intrusion detection system for wireless sensor networks based on a reputation system coupled with kernel self-organizing maps publication-title: Integrated Computer-Aided Engineering – volume: 13 start-page: 1359 year: 2012 end-page: 1368 ident: b33 article-title: An improved feature matching technique for stereo vision applications with the use of self-organizing map publication-title: International Journal of Precision Engineering and Manufacturing – start-page: 58 year: 2011 end-page: 65 ident: b12 article-title: Fast image representation with GPU-based growing neural gas publication-title: Advances in Computational Intelligence – start-page: 155 year: 2008 end-page: 161 ident: b13 article-title: Neural network implementation using cuda and openmp publication-title: Computing: Techniques and applications, 2008. DICTA’08. digital image – start-page: 83 year: 2014 end-page: 100 ident: b26 article-title: A multi-signal variant for the GPU-based parallelization of growing self-organizing networks publication-title: Informatics in control, automation and robotics – year: 2001 ident: b15 article-title: Self-organizing maps, Vol. 30 – start-page: 2533 year: 2009 end-page: 2540 ident: b29 article-title: A growing self-organizing network for reconstructing curves and surfaces publication-title: Neural networks, 2009. IJCNN 2009. international joint conference on – volume: 307 start-page: 14 year: 2018 end-page: 24 ident: b24 article-title: A self-organizing neural network architecture for learning human-object interactions publication-title: Neurocomputing – volume: vol. 341 start-page: 12 year: 2012 end-page: 18 ident: b23 article-title: Scalability of self-organizing maps on a GPU cluster using opencl and CUDA publication-title: Journal of Physics: Conference Series – volume: 44 start-page: 832 year: 2016 end-page: 866 ident: b18 article-title: Mapreduce parallel programming model: A state-of-the-art survey publication-title: International Journal of Parallel Programming – year: 2012 ident: b30 article-title: Self-organizing adaptive map: Autonomous learning of curves and surfaces from point samples publication-title: Neural Networks – start-page: 625 year: 1995 end-page: 632 ident: b8 article-title: A growing neural gas network learns topologies publication-title: Advances in neural information processing systems 7 – start-page: 61 year: 1996 end-page: 72 ident: b9 article-title: Growing self-organizing networks-why publication-title: ESANN, Vol. 96 – volume: 2 year: 2013 ident: b14 article-title: Brain MR image segmentation using self organizing map publication-title: Brain – start-page: 264 year: 2018 end-page: 279 ident: b19 article-title: A scalable heterogeneous parallel SOM based on MPI/CUDA publication-title: Asian conference on machine learning – reference: Qi, C. R., Su, H., Mo, K., & Guibas, L. J. (2017). Pointnet: Deep learning on point sets for 3d classification and segmentation. In – year: 2005 ident: b38 article-title: Self-organizing maps computing on graphic process unit – start-page: 545 year: 2016 end-page: 559 ident: b6 article-title: Mapreduce-based growing neural gas for scalable cluster environments publication-title: Machine learning and data mining in pattern recognition – volume: 31 start-page: 52 year: 2010 end-page: 59 ident: b4 article-title: SOMM: Self organizing Markov map for gesture recognition publication-title: Pattern Recognition Letters – volume: 13 start-page: 68 year: 2011 end-page: 96 ident: b16 article-title: Computational intelligence in wireless sensor networks: A survey publication-title: Communications Surveys & Tutorials, IEEE – start-page: 12 year: 2011 ident: b2 article-title: Cudadma: optimizing GPU memory bandwidth via warp specialization publication-title: Proceedings of 2011 international conference for high performance computing, networking, storage and analysis – volume: 3 start-page: 288 issue: 2 year: 2013 ident: 10.1016/j.neunet.2019.11.016_b28 article-title: Image segmentation by self organizing map with mahalanobis distance publication-title: International Journal of Emerging Technology and Advanced Engineering – volume: 44 start-page: 832 issue: 4 year: 2016 ident: 10.1016/j.neunet.2019.11.016_b18 article-title: Mapreduce parallel programming model: A state-of-the-art survey publication-title: International Journal of Parallel Programming doi: 10.1007/s10766-015-0395-0 – volume: vol. 1 start-page: 622 year: 2005 ident: 10.1016/j.neunet.2019.11.016_b20 article-title: Artificial neural network computation on graphic process unit – start-page: 264 year: 2018 ident: 10.1016/j.neunet.2019.11.016_b19 article-title: A scalable heterogeneous parallel SOM based on MPI/CUDA – start-page: 61 year: 1996 ident: 10.1016/j.neunet.2019.11.016_b9 article-title: Growing self-organizing networks-why – start-page: 625 year: 1995 ident: 10.1016/j.neunet.2019.11.016_b8 article-title: A growing neural gas network learns topologies – volume: 1 start-page: 397 year: 1991 ident: 10.1016/j.neunet.2019.11.016_b22 article-title: A neural-gas network learns topologies publication-title: Artificial Neural Networks – year: 2005 ident: 10.1016/j.neunet.2019.11.016_b38 – volume: 12 start-page: 128 issue: 2 year: 2014 ident: 10.1016/j.neunet.2019.11.016_b34 article-title: GPU-Based optimization of self-organizing map feature matching for real-time stereo vision publication-title: Journal of Information and Communication Convergence Engineering doi: 10.6109/jicce.2014.12.2.128 – start-page: 408 year: 2013 ident: 10.1016/j.neunet.2019.11.016_b36 article-title: Growing neural gas–a parallel approach – start-page: 83 year: 2014 ident: 10.1016/j.neunet.2019.11.016_b26 article-title: A multi-signal variant for the GPU-based parallelization of growing self-organizing networks – volume: 12 issue: 8 year: 2012 ident: 10.1016/j.neunet.2019.11.016_b17 article-title: Integration of growing self-organizing map and continuous genetic algorithm for grading lithium-ion battery cells publication-title: Applied Soft Computing doi: 10.1016/j.asoc.2012.01.018 – start-page: 12 year: 2011 ident: 10.1016/j.neunet.2019.11.016_b2 article-title: Cudadma: optimizing GPU memory bandwidth via warp specialization – volume: 2 issue: 10 year: 2013 ident: 10.1016/j.neunet.2019.11.016_b14 article-title: Brain MR image segmentation using self organizing map publication-title: Brain – start-page: 160 year: 2001 ident: 10.1016/j.neunet.2019.11.016_b32 article-title: Multi-dimensional self-organizing maps on massively parallel hardware – start-page: 63 year: 2017 ident: 10.1016/j.neunet.2019.11.016_b35 article-title: Optimalization of parallel GNG by neurons assigned to processes – start-page: 58 year: 2011 ident: 10.1016/j.neunet.2019.11.016_b12 article-title: Fast image representation with GPU-based growing neural gas publication-title: Advances in Computational Intelligence – volume: vol. 341 start-page: 12 year: 2012 ident: 10.1016/j.neunet.2019.11.016_b23 article-title: Scalability of self-organizing maps on a GPU cluster using opencl and CUDA – volume: 13 start-page: 1359 issue: 8 year: 2012 ident: 10.1016/j.neunet.2019.11.016_b33 article-title: An improved feature matching technique for stereo vision applications with the use of self-organizing map publication-title: International Journal of Precision Engineering and Manufacturing doi: 10.1007/s12541-012-0179-z – volume: 31 start-page: 52 issue: 1 year: 2010 ident: 10.1016/j.neunet.2019.11.016_b4 article-title: SOMM: Self organizing Markov map for gesture recognition publication-title: Pattern Recognition Letters doi: 10.1016/j.patrec.2009.09.009 – volume: 15 start-page: 1041 issue: 8–9 year: 2002 ident: 10.1016/j.neunet.2019.11.016_b21 article-title: A self-organising network that grows when required publication-title: Neural Networks doi: 10.1016/S0893-6080(02)00078-3 – start-page: 1 year: 2011 ident: 10.1016/j.neunet.2019.11.016_b27 article-title: Using self organizing map in wireless sensor network for designing energy efficient topologies – volume: 2005 start-page: 5 year: 2005 ident: 10.1016/j.neunet.2019.11.016_b3 article-title: Graphics hardware implementation of the parameter-less self-organising map publication-title: Intelligent Data Engineering and Automated Learning-IDEAL – volume: 17 start-page: 87 issue: 2 year: 2010 ident: 10.1016/j.neunet.2019.11.016_b1 article-title: Distributed intrusion detection system for wireless sensor networks based on a reputation system coupled with kernel self-organizing maps publication-title: Integrated Computer-Aided Engineering doi: 10.3233/ICA-2010-0334 – start-page: 2533 year: 2009 ident: 10.1016/j.neunet.2019.11.016_b29 article-title: A growing self-organizing network for reconstructing curves and surfaces – start-page: 231 year: 2013 ident: 10.1016/j.neunet.2019.11.016_b11 article-title: GPU Based parallelism for self-organizing map – year: 2001 ident: 10.1016/j.neunet.2019.11.016_b15 – volume: 7 start-page: 1441 issue: 9 year: 1994 ident: 10.1016/j.neunet.2019.11.016_b7 article-title: Growing cell structures–a self-organizing network for unsupervised and supervised learning publication-title: Neural Networks doi: 10.1016/0893-6080(94)90091-4 – ident: 10.1016/j.neunet.2019.11.016_b31 – year: 2012 ident: 10.1016/j.neunet.2019.11.016_b25 article-title: GPGPU implementation of growing neural gas: Application to 3D scene reconstruction publication-title: Journal of Parallel and Distributed Computing doi: 10.1016/j.jpdc.2012.05.008 – start-page: 545 year: 2016 ident: 10.1016/j.neunet.2019.11.016_b6 article-title: Mapreduce-based growing neural gas for scalable cluster environments – year: 2012 ident: 10.1016/j.neunet.2019.11.016_b30 article-title: Self-organizing adaptive map: Autonomous learning of curves and surfaces from point samples publication-title: Neural Networks – volume: 307 start-page: 14 issn: 0925-2312 year: 2018 ident: 10.1016/j.neunet.2019.11.016_b24 article-title: A self-organizing neural network architecture for learning human-object interactions publication-title: Neurocomputing doi: 10.1016/j.neucom.2018.04.015 – volume: 5 start-page: 261 issue: 2 year: 1992 ident: 10.1016/j.neunet.2019.11.016_b37 article-title: Competitive hebbian learning: Algorithm and demonstrations publication-title: Neural Networks doi: 10.1016/S0893-6080(05)80024-3 – volume: 13 start-page: 68 issue: 1 year: 2011 ident: 10.1016/j.neunet.2019.11.016_b16 article-title: Computational intelligence in wireless sensor networks: A survey publication-title: Communications Surveys & Tutorials, IEEE doi: 10.1109/SURV.2011.040310.00002 – start-page: 275 year: 2013 ident: 10.1016/j.neunet.2019.11.016_b10 article-title: Data transfer matters for GPU computing – start-page: 155 year: 2008 ident: 10.1016/j.neunet.2019.11.016_b13 article-title: Neural network implementation using cuda and openmp – start-page: 4591 year: 2010 ident: 10.1016/j.neunet.2019.11.016_b5 article-title: Visual sensing of continuum robot shape using self-organizing maps |
| SSID | ssj0006843 |
| Score | 2.3346403 |
| Snippet | Self-Organizing Neural Networks (SONNs) have a wide range of applications with massive computational requirements that often need to be satisfied with... |
| SourceID | proquest pubmed crossref elsevier |
| SourceType | Aggregation Database Index Database Enrichment Source Publisher |
| StartPage | 108 |
| SubjectTerms | GPU parallelization Self-organizing adaptive map Self-organizing neural networks Surface reconstruction |
| Title | A scalable multi-signal approach for the parallelization of self-organizing neural networks |
| URI | https://dx.doi.org/10.1016/j.neunet.2019.11.016 https://www.ncbi.nlm.nih.gov/pubmed/31838325 https://www.proquest.com/docview/2327381294 |
| Volume | 123 |
| WOSCitedRecordID | wos000511985000009&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVESC databaseName: Elsevier SD Freedom Collection Journals 2021 customDbUrl: eissn: 1879-2782 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0006843 issn: 0893-6080 databaseCode: AIEXJ dateStart: 19950101 isFulltext: true titleUrlDefault: https://www.sciencedirect.com providerName: Elsevier |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1ba9swFBZZu4e97H7JLkWDvQ0FX2RLegylYxsjDNaNwB6MLUvDxbVLnJayn9Bf3SPJcrx0pdtgLyZYUSzO-XJuPheE3tCAS7ADInBLFCWgIQoiwG4mOYtVmeo0UtJy-hNbLPhyKT5PJhe-FuasZk3Dz8_FyX9lNdwDZpvS2b9g9_CjcAM-A9PhCmyH6x8xfv62A7rbiiibLUhMiobpCNB3Dx8SC03X77pWdV-JaczGTtWauEFPP00MwXS7hK2NyxXvxpbs4jdLlnGgU10-_mrdbl5QmflbNgZfgQg-Hhb27Qxju_KtWv2ommqzp8pNGKYvKJLtOD4BzuiQoOWCZr5wZpOlZGWbiEkauCFOM-VkL2eCRIz_KpyjeCRew4CPNHXoqj6vKAEXjziaAY2ACCZ9T8xMp9Zwq-e21eJfzFHMScCSNfOAQZ3vRiwRIOR35x8Olh8HvZ5yl4Ppj-4LMW224NVnXWfoXOfIWIPm8D6623sieO4Q9ABNVPMQ3fNTPnAv9B-h73PsAYXHgMIeUBgAhQFQeAtQuNV4C1DYAQp71DxGX98dHO6_J_1IDiLB0lmTHOzbtKQ8yHPNJGUsLlPweEw-QVFQXuQqlEqEmlJNuRRhXBZBGfA8SiVPTfTsCdpp2kY9QzhIFCjiQiSxLCjTigeSi0LHaaITpZNoimJPv0z2_erN2JQ684mJR5mjemaoDq5sBjeniAy7Tly_lhu-zzxrst7mdLZkBmi6Yedrz8kMRLJ5z5Y3qj3tMnBSGBjCkaBT9NSxeDiLUaGgRJPn__zcF-jO5m_2Eu2sV6fqFbotz9ZVt9pDt9iS7_XQvQTJxb-x |
| linkProvider | Elsevier |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=A+scalable+multi-signal+approach+for+the+parallelization+of+self-organizing+neural+networks&rft.jtitle=Neural+networks&rft.au=Musci%2C+Mirto&rft.au=Parigi%2C+Giacomo&rft.au=Cantoni%2C+Virginio&rft.au=Piastra%2C+Marco&rft.date=2020-03-01&rft.pub=Elsevier+Ltd&rft.issn=0893-6080&rft.eissn=1879-2782&rft.volume=123&rft.spage=108&rft.epage=117&rft_id=info:doi/10.1016%2Fj.neunet.2019.11.016&rft.externalDocID=S0893608019303752 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0893-6080&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0893-6080&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0893-6080&client=summon |