To combat multi-class imbalanced problems by means of over-sampling and boosting techniques
Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties fo...
Ausführliche Beschreibung
Gespeichert in:
| Autor*in: |
Abdi, Lida [verfasserIn] Hashemi, Sattar [verfasserIn] |
|---|
| Format: |
E-Artikel |
|---|---|
| Sprache: |
Englisch |
| Erschienen: |
2014 |
|---|
| Schlagwörter: |
|---|
| Übergeordnetes Werk: |
Enthalten in: Soft Computing - Springer-Verlag, 2003, 19(2014), 12 vom: 30. Apr., Seite 3369-3385 |
|---|---|
| Übergeordnetes Werk: |
volume:19 ; year:2014 ; number:12 ; day:30 ; month:04 ; pages:3369-3385 |
| Links: |
|---|
| DOI / URN: |
10.1007/s00500-014-1291-z |
|---|
| Katalog-ID: |
SPR00648574X |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | SPR00648574X | ||
| 003 | DE-627 | ||
| 005 | 20201124002801.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 201005s2014 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1007/s00500-014-1291-z |2 doi | |
| 035 | |a (DE-627)SPR00648574X | ||
| 035 | |a (SPR)s00500-014-1291-z-e | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Abdi, Lida |e verfasserin |4 aut | |
| 245 | 1 | 3 | |a To combat multi-class imbalanced problems by means of over-sampling and boosting techniques |
| 264 | 1 | |c 2014 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 520 | |a Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties for typical machine learning algorithms. The notion becomes even more complicated when machine learning algorithms are to combat multi-class imbalanced problems. The presented solutions for tackling the issues arising from imbalanced distributions, generally fall into two main categories: data-oriented methods and model-based algorithms. Focusing on the latter, this paper suggests an elegant blend of boosting and over-sampling paradigms, which is called MDOBoost, to bring considerable benefits to the learning ability of multi-class imbalanced data sets. The over-sampling technique introduced and adopted in this paper, Mahalanobis distance-based over-sampling technique (MDO in short), is delicately incorporated into boosting algorithm. In fact, the minority classes are over-sampled via MDO technique in such a way that they almost preserve the original minority class characteristics. MDO, in comparison with the popular method in this field, SMOTE, generates more similar minority class examples to original class samples. Moreover, the broader representation of minority class examples is provided via MDO, and this, in turn, causes the classifier to build larger decision regions. MDOBoost increases the generalization ability of a classifier, since it indicates better results with pruned version of C4.5 classifier; unlike other over-sampling/boosting procedures, which have difficulties with pruned version of C4.5. MDOBoost is applied to real-world multi-class imbalanced benchmarks and its performance is then compared with several data-level and model-based algorithms. The empirical results and theoretical analyses reveal that MDOBoost offers superior advantages compared to popular class decomposition and over-sampling techniques in terms of MAUC, G-mean, and minority class recall. | ||
| 650 | 4 | |a Multi-class imbalance |7 (dpeaa)DE-He213 | |
| 650 | 4 | |a Over-sampling |7 (dpeaa)DE-He213 | |
| 650 | 4 | |a Mahalanobis distance |7 (dpeaa)DE-He213 | |
| 650 | 4 | |a Boosting algorithm |7 (dpeaa)DE-He213 | |
| 650 | 4 | |a Class decomposition techniques |7 (dpeaa)DE-He213 | |
| 700 | 1 | |a Hashemi, Sattar |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Soft Computing |d Springer-Verlag, 2003 |g 19(2014), 12 vom: 30. Apr., Seite 3369-3385 |w (DE-627)SPR006469531 |7 nnns |
| 773 | 1 | 8 | |g volume:19 |g year:2014 |g number:12 |g day:30 |g month:04 |g pages:3369-3385 |
| 856 | 4 | 0 | |u https://dx.doi.org/10.1007/s00500-014-1291-z |z lizenzpflichtig |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a SYSFLAG_A | ||
| 912 | |a GBV_SPRINGER | ||
| 951 | |a AR | ||
| 952 | |d 19 |j 2014 |e 12 |b 30 |c 04 |h 3369-3385 | ||
| author_variant |
l a la s h sh |
|---|---|
| matchkey_str |
abdilidahashemisattar:2014----:oobtutcasmaacdrbesyenooesmln |
| hierarchy_sort_str |
2014 |
| publishDate |
2014 |
| allfields |
10.1007/s00500-014-1291-z doi (DE-627)SPR00648574X (SPR)s00500-014-1291-z-e DE-627 ger DE-627 rakwb eng Abdi, Lida verfasserin aut To combat multi-class imbalanced problems by means of over-sampling and boosting techniques 2014 Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties for typical machine learning algorithms. The notion becomes even more complicated when machine learning algorithms are to combat multi-class imbalanced problems. The presented solutions for tackling the issues arising from imbalanced distributions, generally fall into two main categories: data-oriented methods and model-based algorithms. Focusing on the latter, this paper suggests an elegant blend of boosting and over-sampling paradigms, which is called MDOBoost, to bring considerable benefits to the learning ability of multi-class imbalanced data sets. The over-sampling technique introduced and adopted in this paper, Mahalanobis distance-based over-sampling technique (MDO in short), is delicately incorporated into boosting algorithm. In fact, the minority classes are over-sampled via MDO technique in such a way that they almost preserve the original minority class characteristics. MDO, in comparison with the popular method in this field, SMOTE, generates more similar minority class examples to original class samples. Moreover, the broader representation of minority class examples is provided via MDO, and this, in turn, causes the classifier to build larger decision regions. MDOBoost increases the generalization ability of a classifier, since it indicates better results with pruned version of C4.5 classifier; unlike other over-sampling/boosting procedures, which have difficulties with pruned version of C4.5. MDOBoost is applied to real-world multi-class imbalanced benchmarks and its performance is then compared with several data-level and model-based algorithms. The empirical results and theoretical analyses reveal that MDOBoost offers superior advantages compared to popular class decomposition and over-sampling techniques in terms of MAUC, G-mean, and minority class recall. Multi-class imbalance (dpeaa)DE-He213 Over-sampling (dpeaa)DE-He213 Mahalanobis distance (dpeaa)DE-He213 Boosting algorithm (dpeaa)DE-He213 Class decomposition techniques (dpeaa)DE-He213 Hashemi, Sattar verfasserin aut Enthalten in Soft Computing Springer-Verlag, 2003 19(2014), 12 vom: 30. Apr., Seite 3369-3385 (DE-627)SPR006469531 nnns volume:19 year:2014 number:12 day:30 month:04 pages:3369-3385 https://dx.doi.org/10.1007/s00500-014-1291-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_SPRINGER AR 19 2014 12 30 04 3369-3385 |
| spelling |
10.1007/s00500-014-1291-z doi (DE-627)SPR00648574X (SPR)s00500-014-1291-z-e DE-627 ger DE-627 rakwb eng Abdi, Lida verfasserin aut To combat multi-class imbalanced problems by means of over-sampling and boosting techniques 2014 Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties for typical machine learning algorithms. The notion becomes even more complicated when machine learning algorithms are to combat multi-class imbalanced problems. The presented solutions for tackling the issues arising from imbalanced distributions, generally fall into two main categories: data-oriented methods and model-based algorithms. Focusing on the latter, this paper suggests an elegant blend of boosting and over-sampling paradigms, which is called MDOBoost, to bring considerable benefits to the learning ability of multi-class imbalanced data sets. The over-sampling technique introduced and adopted in this paper, Mahalanobis distance-based over-sampling technique (MDO in short), is delicately incorporated into boosting algorithm. In fact, the minority classes are over-sampled via MDO technique in such a way that they almost preserve the original minority class characteristics. MDO, in comparison with the popular method in this field, SMOTE, generates more similar minority class examples to original class samples. Moreover, the broader representation of minority class examples is provided via MDO, and this, in turn, causes the classifier to build larger decision regions. MDOBoost increases the generalization ability of a classifier, since it indicates better results with pruned version of C4.5 classifier; unlike other over-sampling/boosting procedures, which have difficulties with pruned version of C4.5. MDOBoost is applied to real-world multi-class imbalanced benchmarks and its performance is then compared with several data-level and model-based algorithms. The empirical results and theoretical analyses reveal that MDOBoost offers superior advantages compared to popular class decomposition and over-sampling techniques in terms of MAUC, G-mean, and minority class recall. Multi-class imbalance (dpeaa)DE-He213 Over-sampling (dpeaa)DE-He213 Mahalanobis distance (dpeaa)DE-He213 Boosting algorithm (dpeaa)DE-He213 Class decomposition techniques (dpeaa)DE-He213 Hashemi, Sattar verfasserin aut Enthalten in Soft Computing Springer-Verlag, 2003 19(2014), 12 vom: 30. Apr., Seite 3369-3385 (DE-627)SPR006469531 nnns volume:19 year:2014 number:12 day:30 month:04 pages:3369-3385 https://dx.doi.org/10.1007/s00500-014-1291-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_SPRINGER AR 19 2014 12 30 04 3369-3385 |
| allfields_unstemmed |
10.1007/s00500-014-1291-z doi (DE-627)SPR00648574X (SPR)s00500-014-1291-z-e DE-627 ger DE-627 rakwb eng Abdi, Lida verfasserin aut To combat multi-class imbalanced problems by means of over-sampling and boosting techniques 2014 Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties for typical machine learning algorithms. The notion becomes even more complicated when machine learning algorithms are to combat multi-class imbalanced problems. The presented solutions for tackling the issues arising from imbalanced distributions, generally fall into two main categories: data-oriented methods and model-based algorithms. Focusing on the latter, this paper suggests an elegant blend of boosting and over-sampling paradigms, which is called MDOBoost, to bring considerable benefits to the learning ability of multi-class imbalanced data sets. The over-sampling technique introduced and adopted in this paper, Mahalanobis distance-based over-sampling technique (MDO in short), is delicately incorporated into boosting algorithm. In fact, the minority classes are over-sampled via MDO technique in such a way that they almost preserve the original minority class characteristics. MDO, in comparison with the popular method in this field, SMOTE, generates more similar minority class examples to original class samples. Moreover, the broader representation of minority class examples is provided via MDO, and this, in turn, causes the classifier to build larger decision regions. MDOBoost increases the generalization ability of a classifier, since it indicates better results with pruned version of C4.5 classifier; unlike other over-sampling/boosting procedures, which have difficulties with pruned version of C4.5. MDOBoost is applied to real-world multi-class imbalanced benchmarks and its performance is then compared with several data-level and model-based algorithms. The empirical results and theoretical analyses reveal that MDOBoost offers superior advantages compared to popular class decomposition and over-sampling techniques in terms of MAUC, G-mean, and minority class recall. Multi-class imbalance (dpeaa)DE-He213 Over-sampling (dpeaa)DE-He213 Mahalanobis distance (dpeaa)DE-He213 Boosting algorithm (dpeaa)DE-He213 Class decomposition techniques (dpeaa)DE-He213 Hashemi, Sattar verfasserin aut Enthalten in Soft Computing Springer-Verlag, 2003 19(2014), 12 vom: 30. Apr., Seite 3369-3385 (DE-627)SPR006469531 nnns volume:19 year:2014 number:12 day:30 month:04 pages:3369-3385 https://dx.doi.org/10.1007/s00500-014-1291-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_SPRINGER AR 19 2014 12 30 04 3369-3385 |
| allfieldsGer |
10.1007/s00500-014-1291-z doi (DE-627)SPR00648574X (SPR)s00500-014-1291-z-e DE-627 ger DE-627 rakwb eng Abdi, Lida verfasserin aut To combat multi-class imbalanced problems by means of over-sampling and boosting techniques 2014 Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties for typical machine learning algorithms. The notion becomes even more complicated when machine learning algorithms are to combat multi-class imbalanced problems. The presented solutions for tackling the issues arising from imbalanced distributions, generally fall into two main categories: data-oriented methods and model-based algorithms. Focusing on the latter, this paper suggests an elegant blend of boosting and over-sampling paradigms, which is called MDOBoost, to bring considerable benefits to the learning ability of multi-class imbalanced data sets. The over-sampling technique introduced and adopted in this paper, Mahalanobis distance-based over-sampling technique (MDO in short), is delicately incorporated into boosting algorithm. In fact, the minority classes are over-sampled via MDO technique in such a way that they almost preserve the original minority class characteristics. MDO, in comparison with the popular method in this field, SMOTE, generates more similar minority class examples to original class samples. Moreover, the broader representation of minority class examples is provided via MDO, and this, in turn, causes the classifier to build larger decision regions. MDOBoost increases the generalization ability of a classifier, since it indicates better results with pruned version of C4.5 classifier; unlike other over-sampling/boosting procedures, which have difficulties with pruned version of C4.5. MDOBoost is applied to real-world multi-class imbalanced benchmarks and its performance is then compared with several data-level and model-based algorithms. The empirical results and theoretical analyses reveal that MDOBoost offers superior advantages compared to popular class decomposition and over-sampling techniques in terms of MAUC, G-mean, and minority class recall. Multi-class imbalance (dpeaa)DE-He213 Over-sampling (dpeaa)DE-He213 Mahalanobis distance (dpeaa)DE-He213 Boosting algorithm (dpeaa)DE-He213 Class decomposition techniques (dpeaa)DE-He213 Hashemi, Sattar verfasserin aut Enthalten in Soft Computing Springer-Verlag, 2003 19(2014), 12 vom: 30. Apr., Seite 3369-3385 (DE-627)SPR006469531 nnns volume:19 year:2014 number:12 day:30 month:04 pages:3369-3385 https://dx.doi.org/10.1007/s00500-014-1291-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_SPRINGER AR 19 2014 12 30 04 3369-3385 |
| allfieldsSound |
10.1007/s00500-014-1291-z doi (DE-627)SPR00648574X (SPR)s00500-014-1291-z-e DE-627 ger DE-627 rakwb eng Abdi, Lida verfasserin aut To combat multi-class imbalanced problems by means of over-sampling and boosting techniques 2014 Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties for typical machine learning algorithms. The notion becomes even more complicated when machine learning algorithms are to combat multi-class imbalanced problems. The presented solutions for tackling the issues arising from imbalanced distributions, generally fall into two main categories: data-oriented methods and model-based algorithms. Focusing on the latter, this paper suggests an elegant blend of boosting and over-sampling paradigms, which is called MDOBoost, to bring considerable benefits to the learning ability of multi-class imbalanced data sets. The over-sampling technique introduced and adopted in this paper, Mahalanobis distance-based over-sampling technique (MDO in short), is delicately incorporated into boosting algorithm. In fact, the minority classes are over-sampled via MDO technique in such a way that they almost preserve the original minority class characteristics. MDO, in comparison with the popular method in this field, SMOTE, generates more similar minority class examples to original class samples. Moreover, the broader representation of minority class examples is provided via MDO, and this, in turn, causes the classifier to build larger decision regions. MDOBoost increases the generalization ability of a classifier, since it indicates better results with pruned version of C4.5 classifier; unlike other over-sampling/boosting procedures, which have difficulties with pruned version of C4.5. MDOBoost is applied to real-world multi-class imbalanced benchmarks and its performance is then compared with several data-level and model-based algorithms. The empirical results and theoretical analyses reveal that MDOBoost offers superior advantages compared to popular class decomposition and over-sampling techniques in terms of MAUC, G-mean, and minority class recall. Multi-class imbalance (dpeaa)DE-He213 Over-sampling (dpeaa)DE-He213 Mahalanobis distance (dpeaa)DE-He213 Boosting algorithm (dpeaa)DE-He213 Class decomposition techniques (dpeaa)DE-He213 Hashemi, Sattar verfasserin aut Enthalten in Soft Computing Springer-Verlag, 2003 19(2014), 12 vom: 30. Apr., Seite 3369-3385 (DE-627)SPR006469531 nnns volume:19 year:2014 number:12 day:30 month:04 pages:3369-3385 https://dx.doi.org/10.1007/s00500-014-1291-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_SPRINGER AR 19 2014 12 30 04 3369-3385 |
| language |
English |
| source |
Enthalten in Soft Computing 19(2014), 12 vom: 30. Apr., Seite 3369-3385 volume:19 year:2014 number:12 day:30 month:04 pages:3369-3385 |
| sourceStr |
Enthalten in Soft Computing 19(2014), 12 vom: 30. Apr., Seite 3369-3385 volume:19 year:2014 number:12 day:30 month:04 pages:3369-3385 |
| format_phy_str_mv |
Article |
| institution |
findex.gbv.de |
| topic_facet |
Multi-class imbalance Over-sampling Mahalanobis distance Boosting algorithm Class decomposition techniques |
| isfreeaccess_bool |
false |
| container_title |
Soft Computing |
| authorswithroles_txt_mv |
Abdi, Lida @@aut@@ Hashemi, Sattar @@aut@@ |
| publishDateDaySort_date |
2014-04-30T00:00:00Z |
| hierarchy_top_id |
SPR006469531 |
| id |
SPR00648574X |
| 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">SPR00648574X</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20201124002801.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">201005s2014 xx |||||o 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1007/s00500-014-1291-z</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)SPR00648574X</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(SPR)s00500-014-1291-z-e</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="100" ind1="1" ind2=" "><subfield code="a">Abdi, Lida</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="3"><subfield code="a">To combat multi-class imbalanced problems by means of over-sampling and boosting techniques</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2014</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">Computermedien</subfield><subfield code="b">c</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Online-Ressource</subfield><subfield code="b">cr</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties for typical machine learning algorithms. The notion becomes even more complicated when machine learning algorithms are to combat multi-class imbalanced problems. The presented solutions for tackling the issues arising from imbalanced distributions, generally fall into two main categories: data-oriented methods and model-based algorithms. Focusing on the latter, this paper suggests an elegant blend of boosting and over-sampling paradigms, which is called MDOBoost, to bring considerable benefits to the learning ability of multi-class imbalanced data sets. The over-sampling technique introduced and adopted in this paper, Mahalanobis distance-based over-sampling technique (MDO in short), is delicately incorporated into boosting algorithm. In fact, the minority classes are over-sampled via MDO technique in such a way that they almost preserve the original minority class characteristics. MDO, in comparison with the popular method in this field, SMOTE, generates more similar minority class examples to original class samples. Moreover, the broader representation of minority class examples is provided via MDO, and this, in turn, causes the classifier to build larger decision regions. MDOBoost increases the generalization ability of a classifier, since it indicates better results with pruned version of C4.5 classifier; unlike other over-sampling/boosting procedures, which have difficulties with pruned version of C4.5. MDOBoost is applied to real-world multi-class imbalanced benchmarks and its performance is then compared with several data-level and model-based algorithms. The empirical results and theoretical analyses reveal that MDOBoost offers superior advantages compared to popular class decomposition and over-sampling techniques in terms of MAUC, G-mean, and minority class recall.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Multi-class imbalance</subfield><subfield code="7">(dpeaa)DE-He213</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Over-sampling</subfield><subfield code="7">(dpeaa)DE-He213</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Mahalanobis distance</subfield><subfield code="7">(dpeaa)DE-He213</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Boosting algorithm</subfield><subfield code="7">(dpeaa)DE-He213</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Class decomposition techniques</subfield><subfield code="7">(dpeaa)DE-He213</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Hashemi, Sattar</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Soft Computing</subfield><subfield code="d">Springer-Verlag, 2003</subfield><subfield code="g">19(2014), 12 vom: 30. Apr., Seite 3369-3385</subfield><subfield code="w">(DE-627)SPR006469531</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:19</subfield><subfield code="g">year:2014</subfield><subfield code="g">number:12</subfield><subfield code="g">day:30</subfield><subfield code="g">month:04</subfield><subfield code="g">pages:3369-3385</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://dx.doi.org/10.1007/s00500-014-1291-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_SPRINGER</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">19</subfield><subfield code="j">2014</subfield><subfield code="e">12</subfield><subfield code="b">30</subfield><subfield code="c">04</subfield><subfield code="h">3369-3385</subfield></datafield></record></collection>
|
| author |
Abdi, Lida |
| spellingShingle |
Abdi, Lida misc Multi-class imbalance misc Over-sampling misc Mahalanobis distance misc Boosting algorithm misc Class decomposition techniques To combat multi-class imbalanced problems by means of over-sampling and boosting techniques |
| authorStr |
Abdi, Lida |
| ppnlink_with_tag_str_mv |
@@773@@(DE-627)SPR006469531 |
| format |
electronic Article |
| delete_txt_mv |
keep |
| author_role |
aut aut |
| collection |
springer |
| remote_str |
true |
| illustrated |
Not Illustrated |
| topic_title |
To combat multi-class imbalanced problems by means of over-sampling and boosting techniques Multi-class imbalance (dpeaa)DE-He213 Over-sampling (dpeaa)DE-He213 Mahalanobis distance (dpeaa)DE-He213 Boosting algorithm (dpeaa)DE-He213 Class decomposition techniques (dpeaa)DE-He213 |
| topic |
misc Multi-class imbalance misc Over-sampling misc Mahalanobis distance misc Boosting algorithm misc Class decomposition techniques |
| topic_unstemmed |
misc Multi-class imbalance misc Over-sampling misc Mahalanobis distance misc Boosting algorithm misc Class decomposition techniques |
| topic_browse |
misc Multi-class imbalance misc Over-sampling misc Mahalanobis distance misc Boosting algorithm misc Class decomposition techniques |
| format_facet |
Elektronische Aufsätze Aufsätze Elektronische Ressource |
| format_main_str_mv |
Text Zeitschrift/Artikel |
| carriertype_str_mv |
cr |
| hierarchy_parent_title |
Soft Computing |
| hierarchy_parent_id |
SPR006469531 |
| hierarchy_top_title |
Soft Computing |
| isfreeaccess_txt |
false |
| familylinks_str_mv |
(DE-627)SPR006469531 |
| title |
To combat multi-class imbalanced problems by means of over-sampling and boosting techniques |
| ctrlnum |
(DE-627)SPR00648574X (SPR)s00500-014-1291-z-e |
| title_full |
To combat multi-class imbalanced problems by means of over-sampling and boosting techniques |
| author_sort |
Abdi, Lida |
| journal |
Soft Computing |
| journalStr |
Soft Computing |
| lang_code |
eng |
| isOA_bool |
false |
| recordtype |
marc |
| publishDateSort |
2014 |
| contenttype_str_mv |
txt |
| container_start_page |
3369 |
| author_browse |
Abdi, Lida Hashemi, Sattar |
| container_volume |
19 |
| format_se |
Elektronische Aufsätze |
| author-letter |
Abdi, Lida |
| doi_str_mv |
10.1007/s00500-014-1291-z |
| author2-role |
verfasserin |
| title_sort |
combat multi-class imbalanced problems by means of over-sampling and boosting techniques |
| title_auth |
To combat multi-class imbalanced problems by means of over-sampling and boosting techniques |
| abstract |
Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties for typical machine learning algorithms. The notion becomes even more complicated when machine learning algorithms are to combat multi-class imbalanced problems. The presented solutions for tackling the issues arising from imbalanced distributions, generally fall into two main categories: data-oriented methods and model-based algorithms. Focusing on the latter, this paper suggests an elegant blend of boosting and over-sampling paradigms, which is called MDOBoost, to bring considerable benefits to the learning ability of multi-class imbalanced data sets. The over-sampling technique introduced and adopted in this paper, Mahalanobis distance-based over-sampling technique (MDO in short), is delicately incorporated into boosting algorithm. In fact, the minority classes are over-sampled via MDO technique in such a way that they almost preserve the original minority class characteristics. MDO, in comparison with the popular method in this field, SMOTE, generates more similar minority class examples to original class samples. Moreover, the broader representation of minority class examples is provided via MDO, and this, in turn, causes the classifier to build larger decision regions. MDOBoost increases the generalization ability of a classifier, since it indicates better results with pruned version of C4.5 classifier; unlike other over-sampling/boosting procedures, which have difficulties with pruned version of C4.5. MDOBoost is applied to real-world multi-class imbalanced benchmarks and its performance is then compared with several data-level and model-based algorithms. The empirical results and theoretical analyses reveal that MDOBoost offers superior advantages compared to popular class decomposition and over-sampling techniques in terms of MAUC, G-mean, and minority class recall. |
| abstractGer |
Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties for typical machine learning algorithms. The notion becomes even more complicated when machine learning algorithms are to combat multi-class imbalanced problems. The presented solutions for tackling the issues arising from imbalanced distributions, generally fall into two main categories: data-oriented methods and model-based algorithms. Focusing on the latter, this paper suggests an elegant blend of boosting and over-sampling paradigms, which is called MDOBoost, to bring considerable benefits to the learning ability of multi-class imbalanced data sets. The over-sampling technique introduced and adopted in this paper, Mahalanobis distance-based over-sampling technique (MDO in short), is delicately incorporated into boosting algorithm. In fact, the minority classes are over-sampled via MDO technique in such a way that they almost preserve the original minority class characteristics. MDO, in comparison with the popular method in this field, SMOTE, generates more similar minority class examples to original class samples. Moreover, the broader representation of minority class examples is provided via MDO, and this, in turn, causes the classifier to build larger decision regions. MDOBoost increases the generalization ability of a classifier, since it indicates better results with pruned version of C4.5 classifier; unlike other over-sampling/boosting procedures, which have difficulties with pruned version of C4.5. MDOBoost is applied to real-world multi-class imbalanced benchmarks and its performance is then compared with several data-level and model-based algorithms. The empirical results and theoretical analyses reveal that MDOBoost offers superior advantages compared to popular class decomposition and over-sampling techniques in terms of MAUC, G-mean, and minority class recall. |
| abstract_unstemmed |
Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties for typical machine learning algorithms. The notion becomes even more complicated when machine learning algorithms are to combat multi-class imbalanced problems. The presented solutions for tackling the issues arising from imbalanced distributions, generally fall into two main categories: data-oriented methods and model-based algorithms. Focusing on the latter, this paper suggests an elegant blend of boosting and over-sampling paradigms, which is called MDOBoost, to bring considerable benefits to the learning ability of multi-class imbalanced data sets. The over-sampling technique introduced and adopted in this paper, Mahalanobis distance-based over-sampling technique (MDO in short), is delicately incorporated into boosting algorithm. In fact, the minority classes are over-sampled via MDO technique in such a way that they almost preserve the original minority class characteristics. MDO, in comparison with the popular method in this field, SMOTE, generates more similar minority class examples to original class samples. Moreover, the broader representation of minority class examples is provided via MDO, and this, in turn, causes the classifier to build larger decision regions. MDOBoost increases the generalization ability of a classifier, since it indicates better results with pruned version of C4.5 classifier; unlike other over-sampling/boosting procedures, which have difficulties with pruned version of C4.5. MDOBoost is applied to real-world multi-class imbalanced benchmarks and its performance is then compared with several data-level and model-based algorithms. The empirical results and theoretical analyses reveal that MDOBoost offers superior advantages compared to popular class decomposition and over-sampling techniques in terms of MAUC, G-mean, and minority class recall. |
| collection_details |
GBV_USEFLAG_A SYSFLAG_A GBV_SPRINGER |
| container_issue |
12 |
| title_short |
To combat multi-class imbalanced problems by means of over-sampling and boosting techniques |
| url |
https://dx.doi.org/10.1007/s00500-014-1291-z |
| remote_bool |
true |
| author2 |
Hashemi, Sattar |
| author2Str |
Hashemi, Sattar |
| ppnlink |
SPR006469531 |
| mediatype_str_mv |
c |
| isOA_txt |
false |
| hochschulschrift_bool |
false |
| doi_str |
10.1007/s00500-014-1291-z |
| up_date |
2024-07-03T23:15:00.143Z |
| _version_ |
1803601576088043520 |
| 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">SPR00648574X</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20201124002801.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">201005s2014 xx |||||o 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1007/s00500-014-1291-z</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)SPR00648574X</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(SPR)s00500-014-1291-z-e</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="100" ind1="1" ind2=" "><subfield code="a">Abdi, Lida</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="3"><subfield code="a">To combat multi-class imbalanced problems by means of over-sampling and boosting techniques</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2014</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">Computermedien</subfield><subfield code="b">c</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Online-Ressource</subfield><subfield code="b">cr</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Abstract Imbalanced problems are quite pervasive in many real-world applications. In imbalanced distributions, a class or some classes of data, called minority class(es), is/are under-represented compared to other classes. This skewness in the data underlying distribution causes many difficulties for typical machine learning algorithms. The notion becomes even more complicated when machine learning algorithms are to combat multi-class imbalanced problems. The presented solutions for tackling the issues arising from imbalanced distributions, generally fall into two main categories: data-oriented methods and model-based algorithms. Focusing on the latter, this paper suggests an elegant blend of boosting and over-sampling paradigms, which is called MDOBoost, to bring considerable benefits to the learning ability of multi-class imbalanced data sets. The over-sampling technique introduced and adopted in this paper, Mahalanobis distance-based over-sampling technique (MDO in short), is delicately incorporated into boosting algorithm. In fact, the minority classes are over-sampled via MDO technique in such a way that they almost preserve the original minority class characteristics. MDO, in comparison with the popular method in this field, SMOTE, generates more similar minority class examples to original class samples. Moreover, the broader representation of minority class examples is provided via MDO, and this, in turn, causes the classifier to build larger decision regions. MDOBoost increases the generalization ability of a classifier, since it indicates better results with pruned version of C4.5 classifier; unlike other over-sampling/boosting procedures, which have difficulties with pruned version of C4.5. MDOBoost is applied to real-world multi-class imbalanced benchmarks and its performance is then compared with several data-level and model-based algorithms. The empirical results and theoretical analyses reveal that MDOBoost offers superior advantages compared to popular class decomposition and over-sampling techniques in terms of MAUC, G-mean, and minority class recall.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Multi-class imbalance</subfield><subfield code="7">(dpeaa)DE-He213</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Over-sampling</subfield><subfield code="7">(dpeaa)DE-He213</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Mahalanobis distance</subfield><subfield code="7">(dpeaa)DE-He213</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Boosting algorithm</subfield><subfield code="7">(dpeaa)DE-He213</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Class decomposition techniques</subfield><subfield code="7">(dpeaa)DE-He213</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Hashemi, Sattar</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Soft Computing</subfield><subfield code="d">Springer-Verlag, 2003</subfield><subfield code="g">19(2014), 12 vom: 30. Apr., Seite 3369-3385</subfield><subfield code="w">(DE-627)SPR006469531</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:19</subfield><subfield code="g">year:2014</subfield><subfield code="g">number:12</subfield><subfield code="g">day:30</subfield><subfield code="g">month:04</subfield><subfield code="g">pages:3369-3385</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://dx.doi.org/10.1007/s00500-014-1291-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_SPRINGER</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">19</subfield><subfield code="j">2014</subfield><subfield code="e">12</subfield><subfield code="b">30</subfield><subfield code="c">04</subfield><subfield code="h">3369-3385</subfield></datafield></record></collection>
|
| score |
7.3998833 |