Rapid discrimination and quantification of kudzu root with its adulterant part using FT-NIR and a machine learning algorithm

Two chemometric methods were performed for the determination of acetic acid of fruit vinegars using near infrared (NIR) spectroscopy. Three varieties of fruit vinegars were prepared and 135 samples (45 samples for each variety) were selected for the calibration set, whereas 45 samples (15 samples for each variety) for the validation set. Partial least squares (PLS) analysis was the calibration method as well as extraction method for latent variables (LVs). The first eight LVs were employed as the inputs of least squares-support vector machine (LS-SVM) model. Then LS-SVM model with radial basis function (RBF) kernel was applied to build the regression model compared with PLS model. The correlation coefficient (r), root mean square error of prediction (RMSEP) and bias for validation set were 0.994, 0.814 and -0.091 by PLS, whereas 0.997, 0.651 and 0.011 by LS-SVM, respectively. LS-SVM model outperformed PLS model, but both models achieved an excellent prediction precision. The results indicated that NIR spectroscopy combined with chemometrics could be utilized as a high precision and fast way for the determination of acetic acid of fruit vinegars.

Visible and near infrared (Vis/NIR) spectroscopy combined with a hybrid calibrations of partial least squares (PLS), BP neural network (BPNN) and least squares-support vector machine (LS-SVM) was investigated to determine the pH values of rice vinegars. Five varieties of rice vinegars and 300 samples were prepared. After some preprocessing, PLS was implemented for calibration as well as the extraction of principal components, which would be used as the inputs of BPNN and LS-SVM models. Finally, the BPNN and LS-SVM regression models were developed with a comparison of PLS. The correlation coefficient (r), root mean square error of prediction (RMSEP) and bias for prediction were 0.982, 0.042, and -0.012 by PLS, 0.992, 0.027, and 0.004 by BPNN, while 0.996, 0.018, and 0.005 by LS-SVM, respectively. The results indicated that Vis/NIR spectroscopy combined with chemometrics could be applied as a high precision way for the determination of pH of rice vinegars.

The objective of this paper was to determine soluble solids content (SSC) of intact Gannan navel orange by a portable near-infrared (NIR) spectrometer with the optical fiber in the wavelength range of 551~950nm. The effective wavelength regions (EWs) were chosen from the spectra pro-processed by second derivative by interval partial least square (iPLS) and backward interval partial least square (Bipls). Then the partial least square (PLS) and least square support vector machine (LS-SVM) models were developed with EWs. 60 unknown samples were used to evaluate the performance of them. The LS-SVM model was better than others with EWs chosen by Bilps. The correlation coefficient (R) and root mean square error of prediction (RMSEP) for LS-SVM (Bipls) were 0.86 and 0.55°Brix. The results showed that the portable NIR combination with LS-SVM was a feasible method to determine SSC of intact Gannan navel orange nondestructively.

Discrimination of oils and fuels using a portable NIR spectrometer

Visible/near infrared spectroscopy (Vis/NIRs) technique was applied to non-destructive quantification of sugar and pH value in yogurt. Partial least squares (PLS) analysis and least squares support vector machine (LS-SVM) were implemented for calibration models. In this paper, three brands (Mengniu, Junyao, and Guangming) were set as the calibration, and the remaining two brands (Yili and Shuangfeng) were used as prediction set. In the LS-SVM model, the correlation coefficient (r), root mean square error of prediction, and bias in prediction set were 0.9427, 0.2621°Brix, 1.804e−09 for soluble solids content, and 0.9208, 0.0327, and 1.094e−09 for pH, respectively. The correlation spectra corresponding to the soluble solids content and pH value of yogurt were also analyzed through PLS method. LS-SVM model was better than PLS models for the measurements of soluble solids content and pH value. The results showed that the Vis/NIRs combined with LS-SVM models could predict the soluble solids content and pH value of yogurt.

The present study has aimed at providing new insight into short-wave near-infrared (short-wave NIR) spectroscopy of fat of milk powder. To do that, we analyzed NIR spectra in the 800-1025nm region of 350 milk powder samples. Based on the whole short-wave NIR spectra, performances of least-square support vector machine (LS-SVM) and partial least squares (PLS) are good. Determination coefficients for prediction were up than 0.95, and the root mean square error of prediction (RMSEP) are less than 0.5. The loading weights of PLS and regression coefficients of PLS and LS-SVM were used to determine the sensitive wavelengths for fat content of milk powder. Optimal four sensitive wavelengths, namely 900, 928, 990, and 1018nm, were obtained, and the spectra at these wavelengths were used for the content determination. Rp2 of LS-SVM models are up than 0.97, and RMSEP are less than 0.26. Thus these wavelengths would be useful for the development of portable instrument or online applications to discriminate the fat content of milk powder.

Visible and near infrared (Vis/NIR) spectroscopy combined with least squares-support vector machine (LS-SVM) was investigated for the determination of polysaccharides of auricularia auricula. A total of 240 samples were prepared from four different geographical origins. The calibration set was consisted of 180 samples (45 samples for each origin) and the remaining 60 samples for the validation set. Different preprocessing methods were compared in partial least squares (PLS) models including smoothing, multiplicative scatter correction (MSC), standard normal variate (SNV), the first and second derivative. PLS analysis was employed for the calibration models as well as extraction of certain latent variables (LVs). Simultaneously, some effective wavelengths (EWs) extracted by regression coefficients of LS-SVM were used as the inputs of LS-SVM compared with LVs. The optimal prediction results were achieved by LV-LS-SVM, and the correlation coefficient (r), root mean square error of prediction (RMSEP) and bias for validation set were 0.9413, 0.6893 and -0.0729, respectively. The results were slightly better than PLS. The prediction results of EW-LS-SVM using 3 EWs were acceptable with r=0.9290, RMSEP=0.7714 and Bias=-0.1737. The results indicated that Vis/NIR spectroscopy could be utilized as an efficient way for the determination of polysaccharides of auricularia auricula based on LS-SVM method.

The aim of this study was the non-destructive assessment of ‘Ortanique’ (Citrus reticulata Blanco x Citrus sinensis (L) Osbeck) ripening, based on the prediction of internal quality attributes (IQA) by short-wave near-infrared reflectance spectroscopy (SW-NIRS) calibration models. Spectra from fruit of 50 trees located in two different orchards, were acquired on tree using a customized portable visible near-infrared (vis-NIR) system. Partial least squares (PLS) was used to build the various IQA calibration models. The models were tested through internal validation (IV) and external validation (EV). Generally, the IV results were always superior to those of EV: regarding IV, a high regression coefficient (R2) and low root mean square error of prediction (RMSEP) were achieved, revealing a good predictive performance for juice pH (R2 = 0.80; RMSEP = 0.10; SDR = 2.23), soluble solids content (SSC) (R2 = 0.79; RMSEP = 0.75 %; SDR = 2.27), titratable acidity (TA) (R² = 0.73; RMSEP = 0.24 % citric acid; SDR = 1.94) and the maturation index (MI) (R2 = 0.80; RMSEP = 1.38; SDR = 2.2). The best EV predictions were obtained for TA (R2 = 0.69; RMSEP = 0.38 % citric acid; SDR = 1.24), and MI (R2 = 0.69; RMSEP = 2.07; SDR = 1.49). Calibration models for glucose, fructose and sucrose showed medium-coarse predictions for both validation strategies. A detailed investigation of MI models was performed, to understand the causes of their poor EV results. In the context of EV, model updating strategies were explored by using some validation samples to improve the calibration model. The methods of bias correction and spiking were tested, showing a clear improvement in the predictions.

Detection of adulterated milk using two-dimensional correlation spectroscopy combined with multi-way partial least squares

Online analysis of watercore apples by considering different speeds and orientations based on Vis/NIR full-transmittance spectroscopy

The basic oxygen furnace (BOF) steelmaking process in the steel industry is highly complicated, and subject to variations in raw material composition. During the BOF steelmaking process, it is essential to maintain the carbon content and the endpoint temperature at their set points in the liquid steel. This paper presents intelligent models used to estimate the endpoint temperature in the basic oxygen furnace (BOF) steelmaking process. An artificial neural network (ANN) model and a least-squares support vector machine (LSSVM) model are proposed and their estimation performance compared. The classical partial least-squares (PLS) method was also compared with the others. Results of the estimations using the ANN, LSSVM and PLS models were compared with the operation data, and the root-mean square error (RMSE) for each model was calculated to evaluate estimation performance. The RMSE of the LSSVM model 15.91, which turned out to be the best estimation. RMSE values for the ANN and PLS models were 17.24 and 21.31, respectively, indicating their relative estimation performance. The essential input parameters used in the models can be selected by sensitivity analysis. The RMSE for each model was calculated again after a sequential input selection process was used to remove insignificant input parameters. The RMSE of the LSSVM was then 13.21, which is better than the previous RMSE with all 16 parameters. The results show that LSSVM model using 13 input parameters can be utilized to calculate the required values for oxygen volume and coolant needed to optimally adjust the steel target temperature. (Received July 30, 2018; Accepted September 28, 2018)

The aim of this work is to establish a practical quantitative analysis technique to determine the methanol proportion in methanol–gasoline by using Raman spectroscopy and a hard modeling method. Three hydrocarbon gasoline samples coming from different refineries were added methanol with different concentrations to prepare 27 gasoline–methanol mixture samples with methanol concentration varying from 5% to 70%. Their Raman spectra ranging from 600 to 1800 cm −1 were collected. And then the hard modeling method and two soft methods, namely partial least squares (PLSs) and least squares support vector machine (LSSVM), were performed on the preprocessed spectra to determine methanol concentration. Experimental results show the root mean square error of prediction (RMSEP) based on hard modeling method was 1.02% (v/v), meanwhile that of PLS and LSSVM was 3.28% (v/v) and 3.10% (v/v), respectively. Therefore, the proposed hard modeling method can be particularly applied in complicated mixture samples with overlapping spectra and nonlinear characteristics.

Development of novel univariate and multivariate validated chemometric methods for the analysis of dasatinib, sorafenib, and vandetanib in pure form, dosage forms and biological fluids

This paper tries to explore the optimum variable selection strategy for Partial Least Squares (PLS) regression using a model dataset of cytoprotection data. The compounds of the dataset were classified using K‐means clustering technique applied on standardized descriptor matrix and ten combinations of training and test sets were generated based on the obtained clusters. For a particular training set, PLS models were developed with a number of components optimized by leave‐one‐out Q2 and then the developed models were validated (externally) using the test set compounds. For each set, PLS model was initially constructed using all descriptors (variables). The variables having least standardized values of regression coefficients were deleted and the next model was developed with a reduced set of variables. These steps were performed several times until further reduction in number of variables did not improve Q2 value. In each case, statistical parameters like predictive R2 (R2pred), squared correlation coefficient between observed and predicted values with (r2) and without ($\rm{ r_0^{\rm{2}} }$) intercept and Root Mean Square Error of Prediction (RMSEP) were calculated from the test set compounds. In case of all ten sets, Q2 values steadily increase on deletion of variables while R2pred values do not show any specific trend. In no case, the highest Q2 and highest R2pred appear in the same trial, i.e., with the same combinations of variables. This suggests that from the viewpoint of external predictability, choice of variables for PLS based on Q2 value may not be optimum. Moreover, a clear separation of r2 and r02 curves in some sets suggests that such models may not be truly predictive in spite of acceptable R2pred values. Another observation is that coefficient of determination R2 for the training set is more immune to changes on deletion of variables than the validation parameters like Q2 and R2pred. Finally, a new parameter rm2 has been suggested to indicate external predictability of QSAR models.

ABSTRACTGastrodia elata is a traditional Chinese medicine with medicinal and edible values. In this paper, two kinds of datasets were acquired: partial spectra (artificially obtained peak segment spectra) and full spectra (4000–400 cm−1). Competitive adaptive reweighted sampling algorithm (CARS) and successive projection algorithm (SPA) were utilized to extract the characteristic variables of the two datasets, and Partial Least Squares Discriminant Analysis (PLS‐DA) models, Support Vector Machines (SVM) models, Random Forests (RF) models, and Residual convolutional neural networks (ResNet) were established. It was found that among the PLS‐DA models whole‐MSC‐CARS‐PLS‐DA was optimal, with a Root Mean Square Error of Prediction (RMSEP) of 0.0658; among the SVM models Partial‐Standard Normal Variable (SNV‐SPA‐SVM was the best, with a kernel parameter of 0.1768 and the lowest number of support vectors; among the RF models Partial‐SNV‐RF is optimal, but not as effective as the first two models. The loss value of the ResNet model built based on effective information is 0.001, and the model building time is short and directly uses the original data. Therefore, the ResNet model based on feature bands is the most suitable for practical application compared with other models.