Search

Elastic net (Enet) and sparse partial least squares (SPLS) are frequently employed for wavelength selection and model calibration in analysis of near infrared spectroscopy data. Enet and SPLS can perform variable selection and model calibration simultaneously. And they also tend to select wavelength intervals rather than individual wavelengths when the predictors are multicollinear. In this paper, we focus on comparison of Enet and SPLS in interval wavelength selection and model calibration for near infrared spectroscopy data. The results from both simulation and real spectroscopy data show that Enet method tends to select less predictors as key variables than SPLS; thus it gets more parsimony model and brings advantages for model interpretation. SPLS can obtain much lower mean square of prediction error (MSE) than Enet. So SPLS is more suitable when the attention is to get better model fitting accuracy. The above conclusion is still held when coming to performing the strongly correlated NIR spectroscopy data whose predictors present group structures, Enet exhibits more sparse property than SPLS, and the selected predictors (wavelengths) are segmentally successive. [ABSTRACT FROM AUTHOR]

The partial least squares-discriminant analysis (PLS-DA) is the most widely used statistical tool to perform classification and biomarker screening in metabolomics. However, the PLS-DA tends to overfit the data, and the selection of biomarkers is often unstable because of the disturbance of uninformative variables in principal components. In this paper, we propose an algorithm for performing stable biomarker screening and for seeking the optimal generalization performance, in which the biomarker identification is based on sparse regularization variable selection in combination with subsampling (SRS), and the classification is subsequently performed by a linear support vector machine (SVM) classifier in the selected-variable space to obtain the maximum classification accuracy. Two metabolomics datasets measured by gas chromatography-mass spectrometry are employed to evaluate the performance of the proposed SRS-SVM algorithm, and the comparison with existing related algorithms is given. The result shows that the SRS-SVM algorithm outperforms the PLS-DA and is competitive with other related algorithms in terms of prediction classification accuracy measured by both internal and external validation. Furthermore, the selection of candidate biomarkers is quite stable by the SRS-SVM algorithm, and it can be an alternative and competitive method for the analysis of metabolomics data. The R code for implementing the SRS-SVM algorithm is available in the Electronic supplementary material . [ABSTRACT FROM AUTHOR]

Partial least squares (PLS) is one of the most widely used methods for chemical modeling. However, like many other parameter tunable methods, it has strong tendency of over-fitting. Thus, a crucial step in PLS model building is to select the optimal number of latent variables (nLVs). Cross-validation (CV) is the most popular method for PLS model selection because it selects a model from the perspective of prediction ability. However, a clear minimum of prediction errors may not be obtained in CV which makes the model selection difficult. To solve the problem, we proposed a new strategy for PLS model selection which combines the cross-validated coefficient of determination ( Q c v 2 ) and model stability ( S ). S is defined as the stability of PLS regression vectors which is obtained using model population analysis (MPA). The results show that, when a clear maximum of Q c v 2 is not obtained, S can provide additional information of over-fitting and it helps in finding the optimal nLVs. Compared with other regression vector based indictors such as the Euclidean 2-norm ( B2 ), the Durbin Watson statistic ( DW ) and the jaggedness ( J ), S is more sensitive to over-fitting. The model selected by our method has both good prediction ability and stability. [ABSTRACT FROM AUTHOR]