C-NLPCA: extracting non-linear principal components of image datasets
Autor(a) principal: | |
---|---|
Data de Publicação: | 2005 |
Outros Autores: | , , |
Tipo de documento: | Artigo de conferência |
Idioma: | eng |
Título da fonte: | Repositório Institucional da FURG (RI FURG) |
Texto Completo: | http://repositorio.furg.br/handle/1/1911 |
Resumo: | In this paper we apply a Neural Network (NN)to reduce image dataset, distilling the massive datasets down to a new space of smaller dimension. Due to the possibility of these data have nonlinearities, traditional multi-variate analysis, like the Principal Component Analysis (PCA), may not represent reality. Alternatively, Nonlinear Principal Component Analysis (NLPCA) can be performed by a NN model to fulfill that deficiency. However, when the dimension of the image increases, NN may easily saturate . This work presents an original methodology associated with the use of a set of cascaded multi-layer NN with a bottleneck structure to extract nonlinear information of the large set of image data. We illustrate its good performance with a set of tests against comparisons using this methodology and PCA in the treatment of oceanographic data associated with mesoscale variability of an oceanic boundary current. |
id |
FURG_742727f980528b2777e0e48f920b6290 |
---|---|
oai_identifier_str |
oai:repositorio.furg.br:1/1911 |
network_acronym_str |
FURG |
network_name_str |
Repositório Institucional da FURG (RI FURG) |
repository_id_str |
|
spelling |
Botelho, Silvia Silva da CostaBem, Rodrigo Andrade deAlmeida, Ígor Lorenzato deMata, Mauricio Magalhães2012-03-15T19:56:27Z2012-03-15T19:56:27Z2005BOTELHO, Silvia Silva da Costa, et al. C-NLPCA: extracting non-linear principal components of image datasets. In: 12º Simpósio Brasileiro de Sensoriamento Remoto, 12, Goiânia, 2005. Anais Eletrônicos... Goiânia, 2005. Disponível em:<http://marte.dpi.inpe.br/col/ltid.inpe.br/sbsr/2004/11.22.09.29/doc/3495.pdf>.Acesso em: 15 mar. 2012.http://repositorio.furg.br/handle/1/1911In this paper we apply a Neural Network (NN)to reduce image dataset, distilling the massive datasets down to a new space of smaller dimension. Due to the possibility of these data have nonlinearities, traditional multi-variate analysis, like the Principal Component Analysis (PCA), may not represent reality. Alternatively, Nonlinear Principal Component Analysis (NLPCA) can be performed by a NN model to fulfill that deficiency. However, when the dimension of the image increases, NN may easily saturate . This work presents an original methodology associated with the use of a set of cascaded multi-layer NN with a bottleneck structure to extract nonlinear information of the large set of image data. We illustrate its good performance with a set of tests against comparisons using this methodology and PCA in the treatment of oceanographic data associated with mesoscale variability of an oceanic boundary current.engNeural networkImage processingPCACascaded-NLPCAC-NLPCA: extracting non-linear principal components of image datasetsinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/conferenceObjectinfo:eu-repo/semantics/openAccessreponame:Repositório Institucional da FURG (RI FURG)instname:Universidade Federal do Rio Grande (FURG)instacron:FURGORIGINALC-LPCA.pdfC-LPCA.pdfapplication/pdf319391https://repositorio.furg.br/bitstream/1/1911/1/C-LPCA.pdf549c5dc450e7eb8e6945c27c10448fcfMD51open accessLICENSElicense.txtlicense.txttext/plain; charset=utf-81724https://repositorio.furg.br/bitstream/1/1911/2/license.txt5b92b9704b4f13242d70e45ddef35a68MD52open access1/19112013-02-18 08:20:25.196open accessoai:repositorio.furg.br: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Repositório InstitucionalPUBhttps://repositorio.furg.br/oai/request || http://200.19.254.174/oai/requestopendoar:2013-02-18T11:20:25Repositório Institucional da FURG (RI FURG) - Universidade Federal do Rio Grande (FURG)false |
dc.title.pt_BR.fl_str_mv |
C-NLPCA: extracting non-linear principal components of image datasets |
title |
C-NLPCA: extracting non-linear principal components of image datasets |
spellingShingle |
C-NLPCA: extracting non-linear principal components of image datasets Botelho, Silvia Silva da Costa Neural network Image processing PCA Cascaded-NLPCA |
title_short |
C-NLPCA: extracting non-linear principal components of image datasets |
title_full |
C-NLPCA: extracting non-linear principal components of image datasets |
title_fullStr |
C-NLPCA: extracting non-linear principal components of image datasets |
title_full_unstemmed |
C-NLPCA: extracting non-linear principal components of image datasets |
title_sort |
C-NLPCA: extracting non-linear principal components of image datasets |
author |
Botelho, Silvia Silva da Costa |
author_facet |
Botelho, Silvia Silva da Costa Bem, Rodrigo Andrade de Almeida, Ígor Lorenzato de Mata, Mauricio Magalhães |
author_role |
author |
author2 |
Bem, Rodrigo Andrade de Almeida, Ígor Lorenzato de Mata, Mauricio Magalhães |
author2_role |
author author author |
dc.contributor.author.fl_str_mv |
Botelho, Silvia Silva da Costa Bem, Rodrigo Andrade de Almeida, Ígor Lorenzato de Mata, Mauricio Magalhães |
dc.subject.por.fl_str_mv |
Neural network Image processing PCA Cascaded-NLPCA |
topic |
Neural network Image processing PCA Cascaded-NLPCA |
description |
In this paper we apply a Neural Network (NN)to reduce image dataset, distilling the massive datasets down to a new space of smaller dimension. Due to the possibility of these data have nonlinearities, traditional multi-variate analysis, like the Principal Component Analysis (PCA), may not represent reality. Alternatively, Nonlinear Principal Component Analysis (NLPCA) can be performed by a NN model to fulfill that deficiency. However, when the dimension of the image increases, NN may easily saturate . This work presents an original methodology associated with the use of a set of cascaded multi-layer NN with a bottleneck structure to extract nonlinear information of the large set of image data. We illustrate its good performance with a set of tests against comparisons using this methodology and PCA in the treatment of oceanographic data associated with mesoscale variability of an oceanic boundary current. |
publishDate |
2005 |
dc.date.issued.fl_str_mv |
2005 |
dc.date.accessioned.fl_str_mv |
2012-03-15T19:56:27Z |
dc.date.available.fl_str_mv |
2012-03-15T19:56:27Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/conferenceObject |
format |
conferenceObject |
status_str |
publishedVersion |
dc.identifier.citation.fl_str_mv |
BOTELHO, Silvia Silva da Costa, et al. C-NLPCA: extracting non-linear principal components of image datasets. In: 12º Simpósio Brasileiro de Sensoriamento Remoto, 12, Goiânia, 2005. Anais Eletrônicos... Goiânia, 2005. Disponível em:<http://marte.dpi.inpe.br/col/ltid.inpe.br/sbsr/2004/11.22.09.29/doc/3495.pdf>.Acesso em: 15 mar. 2012. |
dc.identifier.uri.fl_str_mv |
http://repositorio.furg.br/handle/1/1911 |
identifier_str_mv |
BOTELHO, Silvia Silva da Costa, et al. C-NLPCA: extracting non-linear principal components of image datasets. In: 12º Simpósio Brasileiro de Sensoriamento Remoto, 12, Goiânia, 2005. Anais Eletrônicos... Goiânia, 2005. Disponível em:<http://marte.dpi.inpe.br/col/ltid.inpe.br/sbsr/2004/11.22.09.29/doc/3495.pdf>.Acesso em: 15 mar. 2012. |
url |
http://repositorio.furg.br/handle/1/1911 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.source.none.fl_str_mv |
reponame:Repositório Institucional da FURG (RI FURG) instname:Universidade Federal do Rio Grande (FURG) instacron:FURG |
instname_str |
Universidade Federal do Rio Grande (FURG) |
instacron_str |
FURG |
institution |
FURG |
reponame_str |
Repositório Institucional da FURG (RI FURG) |
collection |
Repositório Institucional da FURG (RI FURG) |
bitstream.url.fl_str_mv |
https://repositorio.furg.br/bitstream/1/1911/1/C-LPCA.pdf https://repositorio.furg.br/bitstream/1/1911/2/license.txt |
bitstream.checksum.fl_str_mv |
549c5dc450e7eb8e6945c27c10448fcf 5b92b9704b4f13242d70e45ddef35a68 |
bitstream.checksumAlgorithm.fl_str_mv |
MD5 MD5 |
repository.name.fl_str_mv |
Repositório Institucional da FURG (RI FURG) - Universidade Federal do Rio Grande (FURG) |
repository.mail.fl_str_mv |
|
_version_ |
1798313590116057088 |