A modification of the bootstrapping soft shrinkage approach for spectral variable selection in the issue of over-fitting, model accuracy and variable selection credibility.

In this study, we proposed a new computational method stabilized bootstrapping soft shrinkage approach (SBOSS) for variable selection based on bootstrapping soft shrinkage approach (BOSS) which can enhance the analysis of chemical interest from the massive variables among the overlapped absorption bands. In SBOSS, variable is selected by the index of stability of regression coefficients instead of regression coefficients absolute value. In each loop, a weighted bootstrap sampling (WBS) is applied to generate sub-models, according to the weights update by conducting model population analysis (MPA) on the stability of regression coefficients (RC) of these sub-models. Finally, the subset with the lowest RMSECV is chosen to be the optimal variable set. The performance of the SBOSS was evaluated by one simulated dataset and three NIR datasets. The results show that SBOSS can select the fewer variables and supply the least RMSEP and latent variable number of the PLS model with the best stability comparing with methods of Monte Carlo uninformative variables elimination (MCUVE), genetic algorithm (GA), competitive reweighted sampling (CARS), stability of competitive adaptive reweighted sampling (SCARS) and BOSS.

Geometric search: A new approach for fitting PARAFAC2 models on GC-MS data.

PARAFAC2 is a powerful decomposition method which is ideally suited for modeling gas chromatography-mass spectrometry (GC-MS) data. However, the most widely used fitting algorithms (alternating least squares, ALS) are very slow which hinders use of the model. In this paper, an iterative method called geometric search is proposed to fit the PARAFAC2 model. This method models the PARAFAC2 loading parameters as geometric sequences with offsets during the ALS iterations. It extrapolates the optimal parameters from prior iterations to accelerate ALS convergence process. The performance of this method was evaluated by simulated datasets and two GC-MS datasets of wine and tobacco samples. This geometric search method proved an efficient way to fit PARAFAC2 models, compared with a standard ALS algorithm and two widely used line search algorithms in terms of convergence speed and fitting quality.

Quantitative determination of additive Chlorantraniliprole in Abamectin preparation: Investigation of bootstrapping soft shrinkage approach by mid-infrared spectroscopy.

A novel method, mid-infrared (MIR) spectroscopy, which enables the determination of Chlorantraniliprole in Abamectin within minutes, is proposed. We further evaluate the prediction ability of four wavelength selection methods, including bootstrapping soft shrinkage approach (BOSS), Monte Carlo uninformative variable elimination (MCUVE), genetic algorithm partial least squares (GA-PLS) and competitive adaptive reweighted sampling (CARS) respectively. The results showed that BOSS method obtained the lowest root mean squared error of cross validation (RMSECV) (0.0245) and root mean squared error of prediction (RMSEP) (0.0271), as well as the highest coefficient of determination of cross-validation (Qcv2) (0.9998) and the coefficient of determination of test set (Q2test) (0.9989), which demonstrated that the mid infrared spectroscopy can be used to detect Chlorantraniliprole in Abamectin conveniently. Meanwhile, a suitable wavelength selection method (BOSS) is essential to conducting a component spectral analysis.

[Multivariate Calibration Combined with Mass Spectrometry for Rapid Analysis].

A mixture of four substances of benzaldehyde, iso-octane, butyl acetate, acetophenone were quantitatively analyzed by mass spectrometry combined with chemometrics. The mass chromatogram data of mixture were proceeded with two methods for quantitative analysis. One is feature selection--Multiple Linear Regression (MLR) and the other is full spectrum--Partial Least Squares (PLS). The results show that the RMSEP of benzaldehyde were 0.062 and 0.091 after selecting m/z spectrum and full spectrum respectively; RMSEP of isooctane were 0.048 and 0.057 after selecting spectrum and full spectrum respectively; which of butyl acetate were 0.021 and 0.020 and of acetophenone were 0.010 and 0.032. The feature selection results of the mixture were better than that of the full spectrum modeling results expect butyl acetate which got similar results by the two methods.

Distribution assessment and quantification of counterfeit melamine in powdered milk by NIR imaging methods.

This paper presents a rapid calculation method for the imaging process in the identification and quantification of prohibited additives in milk. Data abstraction methods such as principal component analysis (PCA), classical least squares regression (CLS), and alternative least squares regression (ALS) were used. Different multivariate calculations provided possibilities of quantifying near-infrared (NIR) spectral data cube obtained from the surface of the complex mixture. The results of principal component decomposition confirmed that sample mixture identification is feasible using the PCA-CCI methods. Subsequently, CLSI was used for the direct quantitative analysis of the specific component. Behaving more conveniently than PLS without modeling, CLSI can obtain quantitative information as that melamine generally distribute at the low concentration range of 0-0.5 w/w. Moreover, ALSI can quantify the target component with higher accuracy than CLSI. Standard error of residue to predicted value is 0.0838. Lack of fit is 0.0841. Explanation of variables in the mixture is 99.30%, illustrating that the selective lack of rank is insignificant. Obviously, the most intuitive distribution images are constructed by ALSI among four imaging methods.

A new spectral variable selection pattern using competitive adaptive reweighted sampling combined with successive projections algorithm.

The competitive adaptive reweighted sampling-successive projections algorithm (CARS-SPA) method was proposed as a novel variable selection approach to process multivariate calibration. The CARS was first used to select informative variables, and then SPA to refine the variables with minimum redundant information. The proposed method was applied to near-infrared (NIR) reflectance data of nicotine in tobacco lamina and NIR transmission data of active ingredient in pesticide formulation. As a result, fewer but more informative variables were selected by CARS-SPA than by direct CARS. In the system of pesticide formulation, a multiple linear regression (MLR) model using variables selected by CARS-SPA provided a better prediction than the full-range partial least-squares (PLS) model, successive projections algorithm (SPA) model and uninformative variables elimination-successive projections algorithm (UVE-SPA) processed model. The variable subsets selected by CARS-SPA included the spectral ranges with sufficient chemical information, whereas the uninformative variables were hardly selected.

[Determination of calcium and magnesium in tobacco by near-infrared spectroscopy and least squares-support vector machine].

The purpose of the present paper is to determine calcium and magnesium in tobacco using NIR combined with least squares-support vector machine (LS-SVM). Five hundred ground and dried tobacco samples from Qujing city, Yunnan province, China, were surveyed by a MATRIX-I spectrometer (Bruker Optics, Bremen, Germany). At the beginning of data processing, outliers of samples were eliminated for stability of the model. The rest 487 samples were divided into several calibration sets and validation sets according to a hybrid modeling strategy. Monte-Carlo cross validation was used to choose the best spectral preprocess method from multiplicative scatter correction (MSC), standard normal variate transformation (SNV), S-G smoothing, 1st derivative, etc., and their combinations. To optimize parameters of LS-SVM model, the multilayer grid search and 10-fold cross validation were applied. The final LS-SVM models with the optimizing parameters were trained by the calibration set and accessed by 287 validation samples picked by Kennard-Stone method. For the quantitative model of calcium in tobacco, Savitzky-Golay FIR smoothing with frame size 21 showed the best performance. The regularization parameter λ of LS-SVM was e16.11, while the bandwidth of the RBF kernel σ2 was e8.42. The determination coefficient for prediction (Rc(2)) was 0.9755 and the determination coefficient for prediction (Rp(2)) was 0.9422, better than the performance of PLS model (Rc(2)=0.9593, Rp(2)=0.9344). For the quantitative analysis of magnesium, SNV made the regression model more precise than other preprocess. The optimized λ was e15.25 and σ2 was e6.32. Rc(2) and Rp(2) were 0.9961 and 0.9301, respectively, better than PLS model (Rc(2)=0.9716, Rp(2)=0.8924). After modeling, the whole progress of NIR scan and data analysis for one sample was within tens of seconds. The overall results show that NIR spectroscopy combined with LS-SVM can be efficiently utilized for rapid and accurate analysis of calcium and magnesium in tobacco.

Comparison of several supervised pattern recognition techniques for detecting additive methamidophos in rotenone preparation by near-infrared spectroscopy.

In this paper, different supervised pattern recognition methods have been applied to detect the manually additive methamidophos in rotenone preparation. The aim of this paper was to examine the performances of different supervised pattern recognition techniques: soft independent modeling of class analogy (SIMCA), partial least squares discriminant analysis (PLS-DA), artificial neutral networks (ANN), and support vector machine (SVM). The results obtained show that SVM is the most effective techniques with 100.0% classification accuracy followed by ANN, PLS-DA and with the accuracy of 97.5% and 93.3% respectively while SIMCA yields the poorest result of 85.8%. We hope that the results obtained in this study will help both further chemometric investigations and investigations in the sphere of applied vibrational spectroscopy of sophisticated multicomponent systems. Furthermore, the use of portable instrument and satisfactory classification also indicated the possibility of detecting illicit-addition at scene by near-infrared (NIR) spectroscopy which makes a great sense in pesticide quality control.