Noise reduction for instance-based learning with a local maximal margin approach
Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neigh...
Ausführliche Beschreibung
Gespeichert in:
| Autor*in: |
Segata, Nicola [verfasserIn] |
|---|
| Format: |
Artikel |
|---|---|
| Sprache: |
Englisch |
| Erschienen: |
2009 |
|---|
| Schlagwörter: |
|---|
| Anmerkung: |
© Springer Science+Business Media, LLC 2009 |
|---|
| Übergeordnetes Werk: |
Enthalten in: Journal of intelligent information systems - Springer US, 1992, 35(2009), 2 vom: 20. Aug., Seite 301-331 |
|---|---|
| Übergeordnetes Werk: |
volume:35 ; year:2009 ; number:2 ; day:20 ; month:08 ; pages:301-331 |
| Links: |
|---|
| DOI / URN: |
10.1007/s10844-009-0101-z |
|---|
| Katalog-ID: |
OLC2052417230 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | OLC2052417230 | ||
| 003 | DE-627 | ||
| 005 | 20230503115429.0 | ||
| 007 | tu | ||
| 008 | 200820s2009 xx ||||| 00| ||eng c | ||
| 024 | 7 | |a 10.1007/s10844-009-0101-z |2 doi | |
| 035 | |a (DE-627)OLC2052417230 | ||
| 035 | |a (DE-He213)s10844-009-0101-z-p | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 082 | 0 | 4 | |a 070 |a 020 |a 004 |q VZ |
| 084 | |a 24,1 |2 ssgn | ||
| 084 | |a 54.00 |2 bkl | ||
| 100 | 1 | |a Segata, Nicola |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Noise reduction for instance-based learning with a local maximal margin approach |
| 264 | 1 | |c 2009 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ohne Hilfsmittel zu benutzen |b n |2 rdamedia | ||
| 338 | |a Band |b nc |2 rdacarrier | ||
| 500 | |a © Springer Science+Business Media, LLC 2009 | ||
| 520 | |a Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neighbour classification. There are also other motivations for noise reduction, for instance the elimination of noise may result in simpler models or data cleansing may be an end in itself. In this paper we present a novel approach to noise reduction based on local Support Vector Machines (LSVM) which brings the benefits of maximal margin classifiers to bear on noise reduction. This provides a more robust alternative to the majority rule on which almost all the existing noise reduction techniques are based. Roughly speaking, for each training example an SVM is trained on its neighbourhood and if the SVM classification for the central example disagrees with its actual class there is evidence in favour of removing it from the training set. We provide an empirical evaluation on 15 real datasets showing improved classification accuracy when using training data edited with our method as well as specific experiments regarding the spam filtering application domain. We present a further evaluation on two artificial datasets where we analyse two different types of noise (Gaussian feature noise and mislabelling noise) and the influence of different class densities. The conclusion is that LSVM noise reduction is significantly better than the other analysed algorithms for real datasets and for artificial datasets perturbed by Gaussian noise and in presence of uneven class densities. | ||
| 650 | 4 | |a Noise reduction | |
| 650 | 4 | |a Editing techniques | |
| 650 | 4 | |a -NN | |
| 650 | 4 | |a SVM | |
| 650 | 4 | |a Locality | |
| 700 | 1 | |a Blanzieri, Enrico |4 aut | |
| 700 | 1 | |a Delany, Sarah Jane |4 aut | |
| 700 | 1 | |a Cunningham, Pádraig |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Journal of intelligent information systems |d Springer US, 1992 |g 35(2009), 2 vom: 20. Aug., Seite 301-331 |w (DE-627)171028333 |w (DE-600)1141899-0 |w (DE-576)03304032X |x 0925-9902 |7 nnns |
| 773 | 1 | 8 | |g volume:35 |g year:2009 |g number:2 |g day:20 |g month:08 |g pages:301-331 |
| 856 | 4 | 1 | |u https://doi.org/10.1007/s10844-009-0101-z |z lizenzpflichtig |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a SYSFLAG_A | ||
| 912 | |a GBV_OLC | ||
| 912 | |a SSG-OLC-MAT | ||
| 912 | |a SSG-OLC-BUB | ||
| 912 | |a SSG-OPC-BBI | ||
| 912 | |a GBV_ILN_32 | ||
| 912 | |a GBV_ILN_70 | ||
| 912 | |a GBV_ILN_100 | ||
| 912 | |a GBV_ILN_130 | ||
| 912 | |a GBV_ILN_2006 | ||
| 912 | |a GBV_ILN_2010 | ||
| 912 | |a GBV_ILN_2021 | ||
| 912 | |a GBV_ILN_2244 | ||
| 912 | |a GBV_ILN_4012 | ||
| 912 | |a GBV_ILN_4116 | ||
| 912 | |a GBV_ILN_4324 | ||
| 936 | b | k | |a 54.00 |q VZ |
| 951 | |a AR | ||
| 952 | |d 35 |j 2009 |e 2 |b 20 |c 08 |h 301-331 | ||
| author_variant |
n s ns e b eb s j d sj sjd p c pc |
|---|---|
| matchkey_str |
article:09259902:2009----::osrdcinoisacbsderigihlcla |
| hierarchy_sort_str |
2009 |
| bklnumber |
54.00 |
| publishDate |
2009 |
| allfields |
10.1007/s10844-009-0101-z doi (DE-627)OLC2052417230 (DE-He213)s10844-009-0101-z-p DE-627 ger DE-627 rakwb eng 070 020 004 VZ 24,1 ssgn 54.00 bkl Segata, Nicola verfasserin aut Noise reduction for instance-based learning with a local maximal margin approach 2009 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer Science+Business Media, LLC 2009 Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neighbour classification. There are also other motivations for noise reduction, for instance the elimination of noise may result in simpler models or data cleansing may be an end in itself. In this paper we present a novel approach to noise reduction based on local Support Vector Machines (LSVM) which brings the benefits of maximal margin classifiers to bear on noise reduction. This provides a more robust alternative to the majority rule on which almost all the existing noise reduction techniques are based. Roughly speaking, for each training example an SVM is trained on its neighbourhood and if the SVM classification for the central example disagrees with its actual class there is evidence in favour of removing it from the training set. We provide an empirical evaluation on 15 real datasets showing improved classification accuracy when using training data edited with our method as well as specific experiments regarding the spam filtering application domain. We present a further evaluation on two artificial datasets where we analyse two different types of noise (Gaussian feature noise and mislabelling noise) and the influence of different class densities. The conclusion is that LSVM noise reduction is significantly better than the other analysed algorithms for real datasets and for artificial datasets perturbed by Gaussian noise and in presence of uneven class densities. Noise reduction Editing techniques -NN SVM Locality Blanzieri, Enrico aut Delany, Sarah Jane aut Cunningham, Pádraig aut Enthalten in Journal of intelligent information systems Springer US, 1992 35(2009), 2 vom: 20. Aug., Seite 301-331 (DE-627)171028333 (DE-600)1141899-0 (DE-576)03304032X 0925-9902 nnns volume:35 year:2009 number:2 day:20 month:08 pages:301-331 https://doi.org/10.1007/s10844-009-0101-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT SSG-OLC-BUB SSG-OPC-BBI GBV_ILN_32 GBV_ILN_70 GBV_ILN_100 GBV_ILN_130 GBV_ILN_2006 GBV_ILN_2010 GBV_ILN_2021 GBV_ILN_2244 GBV_ILN_4012 GBV_ILN_4116 GBV_ILN_4324 54.00 VZ AR 35 2009 2 20 08 301-331 |
| spelling |
10.1007/s10844-009-0101-z doi (DE-627)OLC2052417230 (DE-He213)s10844-009-0101-z-p DE-627 ger DE-627 rakwb eng 070 020 004 VZ 24,1 ssgn 54.00 bkl Segata, Nicola verfasserin aut Noise reduction for instance-based learning with a local maximal margin approach 2009 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer Science+Business Media, LLC 2009 Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neighbour classification. There are also other motivations for noise reduction, for instance the elimination of noise may result in simpler models or data cleansing may be an end in itself. In this paper we present a novel approach to noise reduction based on local Support Vector Machines (LSVM) which brings the benefits of maximal margin classifiers to bear on noise reduction. This provides a more robust alternative to the majority rule on which almost all the existing noise reduction techniques are based. Roughly speaking, for each training example an SVM is trained on its neighbourhood and if the SVM classification for the central example disagrees with its actual class there is evidence in favour of removing it from the training set. We provide an empirical evaluation on 15 real datasets showing improved classification accuracy when using training data edited with our method as well as specific experiments regarding the spam filtering application domain. We present a further evaluation on two artificial datasets where we analyse two different types of noise (Gaussian feature noise and mislabelling noise) and the influence of different class densities. The conclusion is that LSVM noise reduction is significantly better than the other analysed algorithms for real datasets and for artificial datasets perturbed by Gaussian noise and in presence of uneven class densities. Noise reduction Editing techniques -NN SVM Locality Blanzieri, Enrico aut Delany, Sarah Jane aut Cunningham, Pádraig aut Enthalten in Journal of intelligent information systems Springer US, 1992 35(2009), 2 vom: 20. Aug., Seite 301-331 (DE-627)171028333 (DE-600)1141899-0 (DE-576)03304032X 0925-9902 nnns volume:35 year:2009 number:2 day:20 month:08 pages:301-331 https://doi.org/10.1007/s10844-009-0101-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT SSG-OLC-BUB SSG-OPC-BBI GBV_ILN_32 GBV_ILN_70 GBV_ILN_100 GBV_ILN_130 GBV_ILN_2006 GBV_ILN_2010 GBV_ILN_2021 GBV_ILN_2244 GBV_ILN_4012 GBV_ILN_4116 GBV_ILN_4324 54.00 VZ AR 35 2009 2 20 08 301-331 |
| allfields_unstemmed |
10.1007/s10844-009-0101-z doi (DE-627)OLC2052417230 (DE-He213)s10844-009-0101-z-p DE-627 ger DE-627 rakwb eng 070 020 004 VZ 24,1 ssgn 54.00 bkl Segata, Nicola verfasserin aut Noise reduction for instance-based learning with a local maximal margin approach 2009 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer Science+Business Media, LLC 2009 Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neighbour classification. There are also other motivations for noise reduction, for instance the elimination of noise may result in simpler models or data cleansing may be an end in itself. In this paper we present a novel approach to noise reduction based on local Support Vector Machines (LSVM) which brings the benefits of maximal margin classifiers to bear on noise reduction. This provides a more robust alternative to the majority rule on which almost all the existing noise reduction techniques are based. Roughly speaking, for each training example an SVM is trained on its neighbourhood and if the SVM classification for the central example disagrees with its actual class there is evidence in favour of removing it from the training set. We provide an empirical evaluation on 15 real datasets showing improved classification accuracy when using training data edited with our method as well as specific experiments regarding the spam filtering application domain. We present a further evaluation on two artificial datasets where we analyse two different types of noise (Gaussian feature noise and mislabelling noise) and the influence of different class densities. The conclusion is that LSVM noise reduction is significantly better than the other analysed algorithms for real datasets and for artificial datasets perturbed by Gaussian noise and in presence of uneven class densities. Noise reduction Editing techniques -NN SVM Locality Blanzieri, Enrico aut Delany, Sarah Jane aut Cunningham, Pádraig aut Enthalten in Journal of intelligent information systems Springer US, 1992 35(2009), 2 vom: 20. Aug., Seite 301-331 (DE-627)171028333 (DE-600)1141899-0 (DE-576)03304032X 0925-9902 nnns volume:35 year:2009 number:2 day:20 month:08 pages:301-331 https://doi.org/10.1007/s10844-009-0101-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT SSG-OLC-BUB SSG-OPC-BBI GBV_ILN_32 GBV_ILN_70 GBV_ILN_100 GBV_ILN_130 GBV_ILN_2006 GBV_ILN_2010 GBV_ILN_2021 GBV_ILN_2244 GBV_ILN_4012 GBV_ILN_4116 GBV_ILN_4324 54.00 VZ AR 35 2009 2 20 08 301-331 |
| allfieldsGer |
10.1007/s10844-009-0101-z doi (DE-627)OLC2052417230 (DE-He213)s10844-009-0101-z-p DE-627 ger DE-627 rakwb eng 070 020 004 VZ 24,1 ssgn 54.00 bkl Segata, Nicola verfasserin aut Noise reduction for instance-based learning with a local maximal margin approach 2009 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer Science+Business Media, LLC 2009 Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neighbour classification. There are also other motivations for noise reduction, for instance the elimination of noise may result in simpler models or data cleansing may be an end in itself. In this paper we present a novel approach to noise reduction based on local Support Vector Machines (LSVM) which brings the benefits of maximal margin classifiers to bear on noise reduction. This provides a more robust alternative to the majority rule on which almost all the existing noise reduction techniques are based. Roughly speaking, for each training example an SVM is trained on its neighbourhood and if the SVM classification for the central example disagrees with its actual class there is evidence in favour of removing it from the training set. We provide an empirical evaluation on 15 real datasets showing improved classification accuracy when using training data edited with our method as well as specific experiments regarding the spam filtering application domain. We present a further evaluation on two artificial datasets where we analyse two different types of noise (Gaussian feature noise and mislabelling noise) and the influence of different class densities. The conclusion is that LSVM noise reduction is significantly better than the other analysed algorithms for real datasets and for artificial datasets perturbed by Gaussian noise and in presence of uneven class densities. Noise reduction Editing techniques -NN SVM Locality Blanzieri, Enrico aut Delany, Sarah Jane aut Cunningham, Pádraig aut Enthalten in Journal of intelligent information systems Springer US, 1992 35(2009), 2 vom: 20. Aug., Seite 301-331 (DE-627)171028333 (DE-600)1141899-0 (DE-576)03304032X 0925-9902 nnns volume:35 year:2009 number:2 day:20 month:08 pages:301-331 https://doi.org/10.1007/s10844-009-0101-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT SSG-OLC-BUB SSG-OPC-BBI GBV_ILN_32 GBV_ILN_70 GBV_ILN_100 GBV_ILN_130 GBV_ILN_2006 GBV_ILN_2010 GBV_ILN_2021 GBV_ILN_2244 GBV_ILN_4012 GBV_ILN_4116 GBV_ILN_4324 54.00 VZ AR 35 2009 2 20 08 301-331 |
| allfieldsSound |
10.1007/s10844-009-0101-z doi (DE-627)OLC2052417230 (DE-He213)s10844-009-0101-z-p DE-627 ger DE-627 rakwb eng 070 020 004 VZ 24,1 ssgn 54.00 bkl Segata, Nicola verfasserin aut Noise reduction for instance-based learning with a local maximal margin approach 2009 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer Science+Business Media, LLC 2009 Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neighbour classification. There are also other motivations for noise reduction, for instance the elimination of noise may result in simpler models or data cleansing may be an end in itself. In this paper we present a novel approach to noise reduction based on local Support Vector Machines (LSVM) which brings the benefits of maximal margin classifiers to bear on noise reduction. This provides a more robust alternative to the majority rule on which almost all the existing noise reduction techniques are based. Roughly speaking, for each training example an SVM is trained on its neighbourhood and if the SVM classification for the central example disagrees with its actual class there is evidence in favour of removing it from the training set. We provide an empirical evaluation on 15 real datasets showing improved classification accuracy when using training data edited with our method as well as specific experiments regarding the spam filtering application domain. We present a further evaluation on two artificial datasets where we analyse two different types of noise (Gaussian feature noise and mislabelling noise) and the influence of different class densities. The conclusion is that LSVM noise reduction is significantly better than the other analysed algorithms for real datasets and for artificial datasets perturbed by Gaussian noise and in presence of uneven class densities. Noise reduction Editing techniques -NN SVM Locality Blanzieri, Enrico aut Delany, Sarah Jane aut Cunningham, Pádraig aut Enthalten in Journal of intelligent information systems Springer US, 1992 35(2009), 2 vom: 20. Aug., Seite 301-331 (DE-627)171028333 (DE-600)1141899-0 (DE-576)03304032X 0925-9902 nnns volume:35 year:2009 number:2 day:20 month:08 pages:301-331 https://doi.org/10.1007/s10844-009-0101-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT SSG-OLC-BUB SSG-OPC-BBI GBV_ILN_32 GBV_ILN_70 GBV_ILN_100 GBV_ILN_130 GBV_ILN_2006 GBV_ILN_2010 GBV_ILN_2021 GBV_ILN_2244 GBV_ILN_4012 GBV_ILN_4116 GBV_ILN_4324 54.00 VZ AR 35 2009 2 20 08 301-331 |
| language |
English |
| source |
Enthalten in Journal of intelligent information systems 35(2009), 2 vom: 20. Aug., Seite 301-331 volume:35 year:2009 number:2 day:20 month:08 pages:301-331 |
| sourceStr |
Enthalten in Journal of intelligent information systems 35(2009), 2 vom: 20. Aug., Seite 301-331 volume:35 year:2009 number:2 day:20 month:08 pages:301-331 |
| format_phy_str_mv |
Article |
| institution |
findex.gbv.de |
| topic_facet |
Noise reduction Editing techniques -NN SVM Locality |
| dewey-raw |
070 |
| isfreeaccess_bool |
false |
| container_title |
Journal of intelligent information systems |
| authorswithroles_txt_mv |
Segata, Nicola @@aut@@ Blanzieri, Enrico @@aut@@ Delany, Sarah Jane @@aut@@ Cunningham, Pádraig @@aut@@ |
| publishDateDaySort_date |
2009-08-20T00:00:00Z |
| hierarchy_top_id |
171028333 |
| dewey-sort |
270 |
| id |
OLC2052417230 |
| language_de |
englisch |
| fullrecord |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">OLC2052417230</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230503115429.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">200820s2009 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1007/s10844-009-0101-z</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC2052417230</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-He213)s10844-009-0101-z-p</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">070</subfield><subfield code="a">020</subfield><subfield code="a">004</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">24,1</subfield><subfield code="2">ssgn</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.00</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Segata, Nicola</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Noise reduction for instance-based learning with a local maximal margin approach</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2009</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="500" ind1=" " ind2=" "><subfield code="a">© Springer Science+Business Media, LLC 2009</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neighbour classification. There are also other motivations for noise reduction, for instance the elimination of noise may result in simpler models or data cleansing may be an end in itself. In this paper we present a novel approach to noise reduction based on local Support Vector Machines (LSVM) which brings the benefits of maximal margin classifiers to bear on noise reduction. This provides a more robust alternative to the majority rule on which almost all the existing noise reduction techniques are based. Roughly speaking, for each training example an SVM is trained on its neighbourhood and if the SVM classification for the central example disagrees with its actual class there is evidence in favour of removing it from the training set. We provide an empirical evaluation on 15 real datasets showing improved classification accuracy when using training data edited with our method as well as specific experiments regarding the spam filtering application domain. We present a further evaluation on two artificial datasets where we analyse two different types of noise (Gaussian feature noise and mislabelling noise) and the influence of different class densities. The conclusion is that LSVM noise reduction is significantly better than the other analysed algorithms for real datasets and for artificial datasets perturbed by Gaussian noise and in presence of uneven class densities.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Noise reduction</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Editing techniques</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">-NN</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">SVM</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Locality</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Blanzieri, Enrico</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Delany, Sarah Jane</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Cunningham, Pádraig</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Journal of intelligent information systems</subfield><subfield code="d">Springer US, 1992</subfield><subfield code="g">35(2009), 2 vom: 20. Aug., Seite 301-331</subfield><subfield code="w">(DE-627)171028333</subfield><subfield code="w">(DE-600)1141899-0</subfield><subfield code="w">(DE-576)03304032X</subfield><subfield code="x">0925-9902</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:35</subfield><subfield code="g">year:2009</subfield><subfield code="g">number:2</subfield><subfield code="g">day:20</subfield><subfield code="g">month:08</subfield><subfield code="g">pages:301-331</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">https://doi.org/10.1007/s10844-009-0101-z</subfield><subfield code="z">lizenzpflichtig</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-MAT</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-BUB</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OPC-BBI</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_32</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_100</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_130</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2006</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2010</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2021</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2244</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4012</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4116</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4324</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.00</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">35</subfield><subfield code="j">2009</subfield><subfield code="e">2</subfield><subfield code="b">20</subfield><subfield code="c">08</subfield><subfield code="h">301-331</subfield></datafield></record></collection>
|
| author |
Segata, Nicola |
| spellingShingle |
Segata, Nicola ddc 070 ssgn 24,1 bkl 54.00 misc Noise reduction misc Editing techniques misc -NN misc SVM misc Locality Noise reduction for instance-based learning with a local maximal margin approach |
| authorStr |
Segata, Nicola |
| ppnlink_with_tag_str_mv |
@@773@@(DE-627)171028333 |
| format |
Article |
| dewey-ones |
070 - News media, journalism & publishing 020 - Library & information sciences 004 - Data processing & computer science |
| delete_txt_mv |
keep |
| author_role |
aut aut aut aut |
| collection |
OLC |
| remote_str |
false |
| illustrated |
Not Illustrated |
| issn |
0925-9902 |
| topic_title |
070 020 004 VZ 24,1 ssgn 54.00 bkl Noise reduction for instance-based learning with a local maximal margin approach Noise reduction Editing techniques -NN SVM Locality |
| topic |
ddc 070 ssgn 24,1 bkl 54.00 misc Noise reduction misc Editing techniques misc -NN misc SVM misc Locality |
| topic_unstemmed |
ddc 070 ssgn 24,1 bkl 54.00 misc Noise reduction misc Editing techniques misc -NN misc SVM misc Locality |
| topic_browse |
ddc 070 ssgn 24,1 bkl 54.00 misc Noise reduction misc Editing techniques misc -NN misc SVM misc Locality |
| format_facet |
Aufsätze Gedruckte Aufsätze |
| format_main_str_mv |
Text Zeitschrift/Artikel |
| carriertype_str_mv |
nc |
| hierarchy_parent_title |
Journal of intelligent information systems |
| hierarchy_parent_id |
171028333 |
| dewey-tens |
070 - News media, journalism & publishing 020 - Library & information sciences 000 - Computer science, knowledge & systems |
| hierarchy_top_title |
Journal of intelligent information systems |
| isfreeaccess_txt |
false |
| familylinks_str_mv |
(DE-627)171028333 (DE-600)1141899-0 (DE-576)03304032X |
| title |
Noise reduction for instance-based learning with a local maximal margin approach |
| ctrlnum |
(DE-627)OLC2052417230 (DE-He213)s10844-009-0101-z-p |
| title_full |
Noise reduction for instance-based learning with a local maximal margin approach |
| author_sort |
Segata, Nicola |
| journal |
Journal of intelligent information systems |
| journalStr |
Journal of intelligent information systems |
| lang_code |
eng |
| isOA_bool |
false |
| dewey-hundreds |
000 - Computer science, information & general works |
| recordtype |
marc |
| publishDateSort |
2009 |
| contenttype_str_mv |
txt |
| container_start_page |
301 |
| author_browse |
Segata, Nicola Blanzieri, Enrico Delany, Sarah Jane Cunningham, Pádraig |
| container_volume |
35 |
| class |
070 020 004 VZ 24,1 ssgn 54.00 bkl |
| format_se |
Aufsätze |
| author-letter |
Segata, Nicola |
| doi_str_mv |
10.1007/s10844-009-0101-z |
| dewey-full |
070 020 004 |
| title_sort |
noise reduction for instance-based learning with a local maximal margin approach |
| title_auth |
Noise reduction for instance-based learning with a local maximal margin approach |
| abstract |
Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neighbour classification. There are also other motivations for noise reduction, for instance the elimination of noise may result in simpler models or data cleansing may be an end in itself. In this paper we present a novel approach to noise reduction based on local Support Vector Machines (LSVM) which brings the benefits of maximal margin classifiers to bear on noise reduction. This provides a more robust alternative to the majority rule on which almost all the existing noise reduction techniques are based. Roughly speaking, for each training example an SVM is trained on its neighbourhood and if the SVM classification for the central example disagrees with its actual class there is evidence in favour of removing it from the training set. We provide an empirical evaluation on 15 real datasets showing improved classification accuracy when using training data edited with our method as well as specific experiments regarding the spam filtering application domain. We present a further evaluation on two artificial datasets where we analyse two different types of noise (Gaussian feature noise and mislabelling noise) and the influence of different class densities. The conclusion is that LSVM noise reduction is significantly better than the other analysed algorithms for real datasets and for artificial datasets perturbed by Gaussian noise and in presence of uneven class densities. © Springer Science+Business Media, LLC 2009 |
| abstractGer |
Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neighbour classification. There are also other motivations for noise reduction, for instance the elimination of noise may result in simpler models or data cleansing may be an end in itself. In this paper we present a novel approach to noise reduction based on local Support Vector Machines (LSVM) which brings the benefits of maximal margin classifiers to bear on noise reduction. This provides a more robust alternative to the majority rule on which almost all the existing noise reduction techniques are based. Roughly speaking, for each training example an SVM is trained on its neighbourhood and if the SVM classification for the central example disagrees with its actual class there is evidence in favour of removing it from the training set. We provide an empirical evaluation on 15 real datasets showing improved classification accuracy when using training data edited with our method as well as specific experiments regarding the spam filtering application domain. We present a further evaluation on two artificial datasets where we analyse two different types of noise (Gaussian feature noise and mislabelling noise) and the influence of different class densities. The conclusion is that LSVM noise reduction is significantly better than the other analysed algorithms for real datasets and for artificial datasets perturbed by Gaussian noise and in presence of uneven class densities. © Springer Science+Business Media, LLC 2009 |
| abstract_unstemmed |
Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neighbour classification. There are also other motivations for noise reduction, for instance the elimination of noise may result in simpler models or data cleansing may be an end in itself. In this paper we present a novel approach to noise reduction based on local Support Vector Machines (LSVM) which brings the benefits of maximal margin classifiers to bear on noise reduction. This provides a more robust alternative to the majority rule on which almost all the existing noise reduction techniques are based. Roughly speaking, for each training example an SVM is trained on its neighbourhood and if the SVM classification for the central example disagrees with its actual class there is evidence in favour of removing it from the training set. We provide an empirical evaluation on 15 real datasets showing improved classification accuracy when using training data edited with our method as well as specific experiments regarding the spam filtering application domain. We present a further evaluation on two artificial datasets where we analyse two different types of noise (Gaussian feature noise and mislabelling noise) and the influence of different class densities. The conclusion is that LSVM noise reduction is significantly better than the other analysed algorithms for real datasets and for artificial datasets perturbed by Gaussian noise and in presence of uneven class densities. © Springer Science+Business Media, LLC 2009 |
| collection_details |
GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT SSG-OLC-BUB SSG-OPC-BBI GBV_ILN_32 GBV_ILN_70 GBV_ILN_100 GBV_ILN_130 GBV_ILN_2006 GBV_ILN_2010 GBV_ILN_2021 GBV_ILN_2244 GBV_ILN_4012 GBV_ILN_4116 GBV_ILN_4324 |
| container_issue |
2 |
| title_short |
Noise reduction for instance-based learning with a local maximal margin approach |
| url |
https://doi.org/10.1007/s10844-009-0101-z |
| remote_bool |
false |
| author2 |
Blanzieri, Enrico Delany, Sarah Jane Cunningham, Pádraig |
| author2Str |
Blanzieri, Enrico Delany, Sarah Jane Cunningham, Pádraig |
| ppnlink |
171028333 |
| mediatype_str_mv |
n |
| isOA_txt |
false |
| hochschulschrift_bool |
false |
| doi_str |
10.1007/s10844-009-0101-z |
| up_date |
2024-07-03T14:59:51.737Z |
| _version_ |
1803570424561270784 |
| fullrecord_marcxml |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">OLC2052417230</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230503115429.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">200820s2009 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1007/s10844-009-0101-z</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC2052417230</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-He213)s10844-009-0101-z-p</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">070</subfield><subfield code="a">020</subfield><subfield code="a">004</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">24,1</subfield><subfield code="2">ssgn</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.00</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Segata, Nicola</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Noise reduction for instance-based learning with a local maximal margin approach</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2009</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="500" ind1=" " ind2=" "><subfield code="a">© Springer Science+Business Media, LLC 2009</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Abstract To some extent the problem of noise reduction in machine learning has been finessed by the development of learning techniques that are noise-tolerant. However, it is difficult to make instance-based learning noise tolerant and noise reduction still plays an important role in k-nearest neighbour classification. There are also other motivations for noise reduction, for instance the elimination of noise may result in simpler models or data cleansing may be an end in itself. In this paper we present a novel approach to noise reduction based on local Support Vector Machines (LSVM) which brings the benefits of maximal margin classifiers to bear on noise reduction. This provides a more robust alternative to the majority rule on which almost all the existing noise reduction techniques are based. Roughly speaking, for each training example an SVM is trained on its neighbourhood and if the SVM classification for the central example disagrees with its actual class there is evidence in favour of removing it from the training set. We provide an empirical evaluation on 15 real datasets showing improved classification accuracy when using training data edited with our method as well as specific experiments regarding the spam filtering application domain. We present a further evaluation on two artificial datasets where we analyse two different types of noise (Gaussian feature noise and mislabelling noise) and the influence of different class densities. The conclusion is that LSVM noise reduction is significantly better than the other analysed algorithms for real datasets and for artificial datasets perturbed by Gaussian noise and in presence of uneven class densities.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Noise reduction</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Editing techniques</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">-NN</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">SVM</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Locality</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Blanzieri, Enrico</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Delany, Sarah Jane</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Cunningham, Pádraig</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Journal of intelligent information systems</subfield><subfield code="d">Springer US, 1992</subfield><subfield code="g">35(2009), 2 vom: 20. Aug., Seite 301-331</subfield><subfield code="w">(DE-627)171028333</subfield><subfield code="w">(DE-600)1141899-0</subfield><subfield code="w">(DE-576)03304032X</subfield><subfield code="x">0925-9902</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:35</subfield><subfield code="g">year:2009</subfield><subfield code="g">number:2</subfield><subfield code="g">day:20</subfield><subfield code="g">month:08</subfield><subfield code="g">pages:301-331</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">https://doi.org/10.1007/s10844-009-0101-z</subfield><subfield code="z">lizenzpflichtig</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-MAT</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-BUB</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OPC-BBI</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_32</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_100</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_130</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2006</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2010</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2021</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2244</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4012</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4116</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4324</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.00</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">35</subfield><subfield code="j">2009</subfield><subfield code="e">2</subfield><subfield code="b">20</subfield><subfield code="c">08</subfield><subfield code="h">301-331</subfield></datafield></record></collection>
|
| score |
7.402895 |