Improved predictions of nonlinear support vector regression and artificial neural network models via preprocessing of data with orthogonal projection to latent structures: A case study
In the presented study, orthogonal projection to latent structures (OPLS) is introduced as a data preprocessing method that handles nonlinear data prior to modelling with two well established nonlinear multivariate models; namely support vector regression (SVR) and artificial neural networks (ANN)....
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Faculty of Pharmacy, Cairo University
2017-12-01
|
| Series: | Bulletin of Faculty of Pharmacy Cairo University |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1110093117300418 |
| _version_ | 1828081431332519936 |
|---|---|
| author | Ibrahim A. Naguib |
| author_facet | Ibrahim A. Naguib |
| author_sort | Ibrahim A. Naguib |
| collection | DOAJ |
| description | In the presented study, orthogonal projection to latent structures (OPLS) is introduced as a data preprocessing method that handles nonlinear data prior to modelling with two well established nonlinear multivariate models; namely support vector regression (SVR) and artificial neural networks (ANN). The proposed preprocessing proved to significantly improve prediction abilities through removal of uncorrelated data.The study was established based on a case study nonlinear spectrofluorimetric data of agomelatine (AGM) and its hydrolysis degradation products (Deg I and Deg II), where a 3 factor 4 level experimental design was used to provide a training set of 16 mixtures with different proportions of studied components. An independent test set which consisted of 9 mixtures was established to confirm the prediction ability of the introduced models. Excitation wavelength was 227 nm, and working range for emission spectra was 320â440 nm.The couplings of OPLS-SVR and OPLS-ANN provided better accuracy for prediction of independent nonlinear test set. The root mean square error of prediction RMSEP for the test set mixtures was used as a major comparison parameter, where RMSEP results for OPLS-SVR and OPLS-ANN are 2.19 and 1.50 respectively. Keywords: Agomelatine, SVR, ANN, OPLS, Spectrofluorimetry, Nonlinear |
| first_indexed | 2024-04-11T03:34:58Z |
| format | Article |
| id | doaj.art-d555868d01b44cdfadcba98c8b82764d |
| institution | Directory Open Access Journal |
| issn | 1110-0931 |
| language | English |
| last_indexed | 2024-04-11T03:34:58Z |
| publishDate | 2017-12-01 |
| publisher | Faculty of Pharmacy, Cairo University |
| record_format | Article |
| series | Bulletin of Faculty of Pharmacy Cairo University |
| spelling | doaj.art-d555868d01b44cdfadcba98c8b82764d2023-01-02T05:26:50ZengFaculty of Pharmacy, Cairo UniversityBulletin of Faculty of Pharmacy Cairo University1110-09312017-12-01552287291Improved predictions of nonlinear support vector regression and artificial neural network models via preprocessing of data with orthogonal projection to latent structures: A case studyIbrahim A. Naguib0Address: Pharmaceutical Analytical Chemistry Department, Faculty of Pharmacy, Beni-Suef University, Alshaheed Shehata Ahmad Hegazy St., 62514 Beni-Suef, Egypt.; Pharmaceutical Analytical Chemistry Department, Faculty of Pharmacy, Beni-Suef University, Alshaheed Shehata Ahmad Hegazy St., 62514 Beni-Suef, Egypt; Pharmaceutical Chemistry Department, Faculty of Pharmacy, University of Tabuk, Tabuk, Saudi ArabiaIn the presented study, orthogonal projection to latent structures (OPLS) is introduced as a data preprocessing method that handles nonlinear data prior to modelling with two well established nonlinear multivariate models; namely support vector regression (SVR) and artificial neural networks (ANN). The proposed preprocessing proved to significantly improve prediction abilities through removal of uncorrelated data.The study was established based on a case study nonlinear spectrofluorimetric data of agomelatine (AGM) and its hydrolysis degradation products (Deg I and Deg II), where a 3 factor 4 level experimental design was used to provide a training set of 16 mixtures with different proportions of studied components. An independent test set which consisted of 9 mixtures was established to confirm the prediction ability of the introduced models. Excitation wavelength was 227 nm, and working range for emission spectra was 320â440 nm.The couplings of OPLS-SVR and OPLS-ANN provided better accuracy for prediction of independent nonlinear test set. The root mean square error of prediction RMSEP for the test set mixtures was used as a major comparison parameter, where RMSEP results for OPLS-SVR and OPLS-ANN are 2.19 and 1.50 respectively. Keywords: Agomelatine, SVR, ANN, OPLS, Spectrofluorimetry, Nonlinearhttp://www.sciencedirect.com/science/article/pii/S1110093117300418 |
| spellingShingle | Ibrahim A. Naguib Improved predictions of nonlinear support vector regression and artificial neural network models via preprocessing of data with orthogonal projection to latent structures: A case study Bulletin of Faculty of Pharmacy Cairo University |
| title | Improved predictions of nonlinear support vector regression and artificial neural network models via preprocessing of data with orthogonal projection to latent structures: A case study |
| title_full | Improved predictions of nonlinear support vector regression and artificial neural network models via preprocessing of data with orthogonal projection to latent structures: A case study |
| title_fullStr | Improved predictions of nonlinear support vector regression and artificial neural network models via preprocessing of data with orthogonal projection to latent structures: A case study |
| title_full_unstemmed | Improved predictions of nonlinear support vector regression and artificial neural network models via preprocessing of data with orthogonal projection to latent structures: A case study |
| title_short | Improved predictions of nonlinear support vector regression and artificial neural network models via preprocessing of data with orthogonal projection to latent structures: A case study |
| title_sort | improved predictions of nonlinear support vector regression and artificial neural network models via preprocessing of data with orthogonal projection to latent structures a case study |
| url | http://www.sciencedirect.com/science/article/pii/S1110093117300418 |
| work_keys_str_mv | AT ibrahimanaguib improvedpredictionsofnonlinearsupportvectorregressionandartificialneuralnetworkmodelsviapreprocessingofdatawithorthogonalprojectiontolatentstructuresacasestudy |