So you think you can PLS-DA?
Abstract Background Partial Least-Squares Discriminant Analysis (PLS-DA) is a popular machine learning tool that is gaining increasing attention as a useful feature selector and classifier. In an effort to understand its strengths and weaknesses, we performed a series of experiments with synthetic d...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2020-12-01
|
| Series: | BMC Bioinformatics |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12859-019-3310-7 |
| _version_ | 1818623551999049728 |
|---|---|
| author | Daniel Ruiz-Perez Haibin Guan Purnima Madhivanan Kalai Mathee Giri Narasimhan |
| author_facet | Daniel Ruiz-Perez Haibin Guan Purnima Madhivanan Kalai Mathee Giri Narasimhan |
| author_sort | Daniel Ruiz-Perez |
| collection | DOAJ |
| description | Abstract Background Partial Least-Squares Discriminant Analysis (PLS-DA) is a popular machine learning tool that is gaining increasing attention as a useful feature selector and classifier. In an effort to understand its strengths and weaknesses, we performed a series of experiments with synthetic data and compared its performance to its close relative from which it was initially invented, namely Principal Component Analysis (PCA). Results We demonstrate that even though PCA ignores the information regarding the class labels of the samples, this unsupervised tool can be remarkably effective as a feature selector. In some cases, it outperforms PLS-DA, which is made aware of the class labels in its input. Our experiments range from looking at the signal-to-noise ratio in the feature selection task, to considering many practical distributions and models encountered when analyzing bioinformatics and clinical data. Other methods were also evaluated. Finally, we analyzed an interesting data set from 396 vaginal microbiome samples where the ground truth for the feature selection was available. All the 3D figures shown in this paper as well as the supplementary ones can be viewed interactively at http://biorg.cs.fiu.edu/plsda Conclusions Our results highlighted the strengths and weaknesses of PLS-DA in comparison with PCA for different underlying data models. |
| first_indexed | 2024-12-16T18:42:52Z |
| format | Article |
| id | doaj.art-6a907005b0d448f7adfb27c5cfbdf021 |
| institution | Directory Open Access Journal |
| issn | 1471-2105 |
| language | English |
| last_indexed | 2024-12-16T18:42:52Z |
| publishDate | 2020-12-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Bioinformatics |
| spelling | doaj.art-6a907005b0d448f7adfb27c5cfbdf0212022-12-21T22:20:58ZengBMCBMC Bioinformatics1471-21052020-12-0121S111010.1186/s12859-019-3310-7So you think you can PLS-DA?Daniel Ruiz-Perez0Haibin Guan1Purnima Madhivanan2Kalai Mathee3Giri Narasimhan4Bioinformatics Research Group (BioRG), Florida International UniversityBioinformatics Research Group (BioRG), Florida International UniversityDepartment of Epidemiology, Florida International UniversityHerbert Wertheim College of Medicine, Florida International UniversityBioinformatics Research Group (BioRG), Florida International UniversityAbstract Background Partial Least-Squares Discriminant Analysis (PLS-DA) is a popular machine learning tool that is gaining increasing attention as a useful feature selector and classifier. In an effort to understand its strengths and weaknesses, we performed a series of experiments with synthetic data and compared its performance to its close relative from which it was initially invented, namely Principal Component Analysis (PCA). Results We demonstrate that even though PCA ignores the information regarding the class labels of the samples, this unsupervised tool can be remarkably effective as a feature selector. In some cases, it outperforms PLS-DA, which is made aware of the class labels in its input. Our experiments range from looking at the signal-to-noise ratio in the feature selection task, to considering many practical distributions and models encountered when analyzing bioinformatics and clinical data. Other methods were also evaluated. Finally, we analyzed an interesting data set from 396 vaginal microbiome samples where the ground truth for the feature selection was available. All the 3D figures shown in this paper as well as the supplementary ones can be viewed interactively at http://biorg.cs.fiu.edu/plsda Conclusions Our results highlighted the strengths and weaknesses of PLS-DA in comparison with PCA for different underlying data models.https://doi.org/10.1186/s12859-019-3310-7PLS-DAPCAFeature selectionDimensionality reductionBioinformatics |
| spellingShingle | Daniel Ruiz-Perez Haibin Guan Purnima Madhivanan Kalai Mathee Giri Narasimhan So you think you can PLS-DA? BMC Bioinformatics PLS-DA PCA Feature selection Dimensionality reduction Bioinformatics |
| title | So you think you can PLS-DA? |
| title_full | So you think you can PLS-DA? |
| title_fullStr | So you think you can PLS-DA? |
| title_full_unstemmed | So you think you can PLS-DA? |
| title_short | So you think you can PLS-DA? |
| title_sort | so you think you can pls da |
| topic | PLS-DA PCA Feature selection Dimensionality reduction Bioinformatics |
| url | https://doi.org/10.1186/s12859-019-3310-7 |
| work_keys_str_mv | AT danielruizperez soyouthinkyoucanplsda AT haibinguan soyouthinkyoucanplsda AT purnimamadhivanan soyouthinkyoucanplsda AT kalaimathee soyouthinkyoucanplsda AT girinarasimhan soyouthinkyoucanplsda |