Effects of Pooling Samples on the Performance of Classification Algorithms: A Comparative Study
A pooling design can be used as a powerful strategy to compensate for limited amounts of samples or high biological variation. In this paper, we perform a comparative study to model and quantify the effects of virtual pooling on the performance of the widely applied classifiers, support vector machi...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2012-01-01
|
| Series: | The Scientific World Journal |
| Online Access: | http://dx.doi.org/10.1100/2012/278352 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832545763655155712 |
|---|---|
| author | Kanthida Kusonmano Michael Netzer Christian Baumgartner Matthias Dehmer Klaus R. Liedl Armin Graber |
| author_facet | Kanthida Kusonmano Michael Netzer Christian Baumgartner Matthias Dehmer Klaus R. Liedl Armin Graber |
| author_sort | Kanthida Kusonmano |
| collection | DOAJ |
| description | A pooling design can be used as a powerful strategy to compensate for limited amounts of samples or high biological variation. In this paper, we perform a comparative study to model and quantify the effects of virtual pooling on the performance of the widely applied classifiers, support vector machines (SVMs), random forest (RF), k-nearest neighbors (k-NN), penalized logistic regression (PLR), and prediction analysis for microarrays (PAMs). We evaluate a variety of experimental designs using mock omics datasets with varying levels of pool sizes and considering effects from feature selection. Our results show that feature selection significantly improves classifier performance for non-pooled and pooled data. All investigated classifiers yield lower misclassification rates with smaller pool sizes. RF mainly outperforms other investigated algorithms, while accuracy levels are comparable among all the remaining ones. Guidelines are derived to identify an optimal pooling scheme for obtaining adequate predictive power and, hence, to motivate a study design that meets best experimental objectives and budgetary conditions, including time constraints. |
| format | Article |
| id | doaj-art-21307f9d6ba94534911143bde69cdf57 |
| institution | Kabale University |
| issn | 1537-744X |
| language | English |
| publishDate | 2012-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | The Scientific World Journal |
| spelling | doaj-art-21307f9d6ba94534911143bde69cdf572025-02-03T07:24:47ZengWileyThe Scientific World Journal1537-744X2012-01-01201210.1100/2012/278352278352Effects of Pooling Samples on the Performance of Classification Algorithms: A Comparative StudyKanthida Kusonmano0Michael Netzer1Christian Baumgartner2Matthias Dehmer3Klaus R. Liedl4Armin Graber5Institute for Bioinformatics and Translational Research, UMIT, 6060 Hall in Tyrol, AustriaInstitute of Electrical and Biomedical Engineering, UMIT, 6060 Hall in Tyrol, AustriaInstitute of Electrical and Biomedical Engineering, UMIT, 6060 Hall in Tyrol, AustriaInstitute for Bioinformatics and Translational Research, UMIT, 6060 Hall in Tyrol, AustriaFaculty of Chemistry and Pharmacy, Leopold-Franzens-University Innsbruck, 6020 Innsbruck, AustriaInstitute for Bioinformatics and Translational Research, UMIT, 6060 Hall in Tyrol, AustriaA pooling design can be used as a powerful strategy to compensate for limited amounts of samples or high biological variation. In this paper, we perform a comparative study to model and quantify the effects of virtual pooling on the performance of the widely applied classifiers, support vector machines (SVMs), random forest (RF), k-nearest neighbors (k-NN), penalized logistic regression (PLR), and prediction analysis for microarrays (PAMs). We evaluate a variety of experimental designs using mock omics datasets with varying levels of pool sizes and considering effects from feature selection. Our results show that feature selection significantly improves classifier performance for non-pooled and pooled data. All investigated classifiers yield lower misclassification rates with smaller pool sizes. RF mainly outperforms other investigated algorithms, while accuracy levels are comparable among all the remaining ones. Guidelines are derived to identify an optimal pooling scheme for obtaining adequate predictive power and, hence, to motivate a study design that meets best experimental objectives and budgetary conditions, including time constraints.http://dx.doi.org/10.1100/2012/278352 |
| spellingShingle | Kanthida Kusonmano Michael Netzer Christian Baumgartner Matthias Dehmer Klaus R. Liedl Armin Graber Effects of Pooling Samples on the Performance of Classification Algorithms: A Comparative Study The Scientific World Journal |
| title | Effects of Pooling Samples on the Performance of Classification Algorithms: A Comparative Study |
| title_full | Effects of Pooling Samples on the Performance of Classification Algorithms: A Comparative Study |
| title_fullStr | Effects of Pooling Samples on the Performance of Classification Algorithms: A Comparative Study |
| title_full_unstemmed | Effects of Pooling Samples on the Performance of Classification Algorithms: A Comparative Study |
| title_short | Effects of Pooling Samples on the Performance of Classification Algorithms: A Comparative Study |
| title_sort | effects of pooling samples on the performance of classification algorithms a comparative study |
| url | http://dx.doi.org/10.1100/2012/278352 |
| work_keys_str_mv | AT kanthidakusonmano effectsofpoolingsamplesontheperformanceofclassificationalgorithmsacomparativestudy AT michaelnetzer effectsofpoolingsamplesontheperformanceofclassificationalgorithmsacomparativestudy AT christianbaumgartner effectsofpoolingsamplesontheperformanceofclassificationalgorithmsacomparativestudy AT matthiasdehmer effectsofpoolingsamplesontheperformanceofclassificationalgorithmsacomparativestudy AT klausrliedl effectsofpoolingsamplesontheperformanceofclassificationalgorithmsacomparativestudy AT armingraber effectsofpoolingsamplesontheperformanceofclassificationalgorithmsacomparativestudy |