Rapid detection of three quality parameters and classification of wine based on Vis-NIR spectroscopy with wavelength selection by ACO and CARS algorithms

Quantitative analysis of three ingredients in Salvia miltiorrhiza by near infrared spectroscopy combined with hybrid variable selection strategy

Rosmarinic acid (RA), Tanshinone IIA (Tan IIA), and Salvianolic acid B (Sal B) are crucial compounds found in Salvia miltiorrhiza. Quickly predicting these components can aid in ensuring the quality of S. miltiorrhiza. Spectral preprocessing and variable selection are essential processes in quantitative analysis using near infrared spectroscopy (NIR). A novel hybrid variable selection approach utilizing iVISSA was employed in this study to enhance the quantitative measurement of RA, Tan IIA, and Sal B contents in S. miltiorrhiza. The spectra underwent 108 preprocessing approaches, with the optimal method being determined as orthogonal signal correction (OSC). iVISSA was utilized to identify the intervals (feature bands) that were most pertinent to the target chemical. Various methods such as bootstrapping soft shrinkage (BOSS), competitive adaptive reweighted sampling (CARS), genetic algorithm (GA), variable combination population analysis (VCPA), successive projections algorithm (SPA), iteratively variable subset optimization (IVSO), and iteratively retained informative variables (IRIV) were used to identify significant feature variables. PLSR models were created for comparison using the given variables. The results fully demonstrated that iVISSA-SPA calibration model had the best comprehensive performance for Tan IIA, and iVISSA-BOSS had the best comprehensive performance for RA and Sal B, and correlation coefficients of cross-validation (R2cv), root mean square errors of cross-validation (RMSECV), correlation coefficients of prediction (R2p), and root mean square errors of prediction (RMSEP) were 0.9970, 0.0054, 0.9990 and 0.0033, 0.9992, 0.0016, 0.9961 and 0.0034, 0.9998, 0.0138, 0.9875 and 0.1090, respectively. The results suggest that NIR spectroscopy, along with PLSR and a hybrid variable selection method using iVISSA, can be a valuable tool for quickly quantifying RA, Sal B, and Tan IIA in S. miltiorrhiza.

Application of terahertz Time-Domain spectroscopy and chemometrics-based whale optimization algorithm in PDE5 inhibitor detection

Combating the illicit use of PDE5 inhibitor drugs is a focal point in forensic science research. In order to achieve rapid identification of such drugs, this study applies terahertz time-domain spectroscopy combined with chemometrics to establish a fast and accurate detection method for PDE5 inhibitors. The optimal detection method is determined by comparing the spectral performance of three optical parameters, namely absorption coefficient, refractive index, and dielectric constant. Linear discriminant models based on different spectral parameters, whale optimization algorithm optimized extreme learning machine models, and whale optimization algorithm optimized random forest models are established. The effectiveness and performance of principal component analysis and competitive adaptive reweighted sampling algorithm for spectral feature data selection are also investigated. The PDE5 inhibitor identification model based on the competitive adaptive reweighted sampling - whale optimization algorithm - random forest (CARS-WOA-RF) model achieves an accuracy of 98.61%, and the identification model for two concentrations of Sildenafil achieves 100% accuracy. The results demonstrate that terahertz time-domain spectroscopy combined with chemometrics can effectively detect various common types of PDE5 inhibitor drugs and different concentrations.

Improvement of the prediction of a visual apple ripeness index under seasonal variation by NIR spectral model correction

Apple ripeness assessment is essential to ensure its post-harvest commercial value, and the visible/near-infrared(NIR) spectral models that are effective in achieving this goal are prone to failure due to seasonal or instrumental factors. This study has proposed a visual ripeness index (VRPI) determined by parameters such as soluble solids, titratable acids, etc., which vary during the ripening period of the apple. The R and RMSE of the index prediction model based on the 2019 sample were 0.871 to 0.913 and 0.184 to 0.213 respectively. The model failed to predict the next two years of the sample, which was effectively addressed by model fusion and correction. For the 2020 and 2021 samples, the revised model improves R by 6.8% and 10.6% and reduces RMSE by 52.2% and 32.2% respectively. The results showed that the global model is adapted to the correction of the VRPI spectral prediction model under seasonal variation.

Application of colorimetric sensor array coupled with machine-learning approaches for the discrimination of grains based on freshness

BACKGROUND

Volatile organic compounds (VOCs) in grain fluctuate depending on the degree of grain freshness. A new colorimetric sensor array (CSA) was developed as capture probes for the quantification of VOCs in grains in this work, and it was designed to monitor the variation of grain VOCs. CSA spectral data acquisition using visible-near-infrared spectroscopy and image processing of CSA's image imformation by computer were used comparatively. Then, machine-learning-based models - for example, synergistic interval partial least squares, genetic algorithm, competitive adaptive reweighted sampling (CARS) algorithm, and ant colony optimization (ACO) algorithm - were introduced to optimize variables. Moreover, principal component analysis, and linear discriminant analysis (LDA), and K-nearest neighbors (KNN) were used for the classification. Ultimately, quantitative models for detecting grain freshness are developed using various variable selection strategies.

RESULTS

Compared with the pattern recognition results of image processing, visible-near-infrared spectroscopy could better separate the grains with different freshness from principal component analysis, and the prediction set of LDA models could correctly identify 100% of rice, 96.88% of paddy, and 97.9% of soybeans. In addition, compared with CARS and ACO, the LDA model and KNN model based on genetic algorithms show the best prediction performance. The prediction set could correctly identify 100% of rice and paddy samples and 95.83% of soybean samples.

CONCLUSION

The method developed could be used for non-destructive detection of grain freshness. © 2023 Society of Chemical Industry.

Spoilage Monitoring and Early Warning for Apples in Storage Using Gas Sensors and Chemometrics

In the process of storage and cold chain logistics, apples are prone to physical bumps or microbial infection, which easily leads to spoilage in the micro-environment, resulting in widespread infection and serious post-harvest economic losses. Thus, development of methods for monitoring apple spoilage and providing early warning of spoilage has become the focus for post-harvest loss reduction. Thus, in this study, a spoilage monitoring and early warning system was developed by measuring volatile component production during apple spoilage combined with chemometric analysis. An apple spoilage monitoring prototype was designed to include a gas monitoring array capable of measuring volatile organic compounds, such as CO2, O2 and C2H4, integrated with the temperature and humidity sensor. The sensor information from a simulated apple warehouse was obtained by the prototype, and a multi-factor fusion early warning model of apple spoilage was established based on various modeling methods. Simulated annealing-partial least squares (SA-PLS) was the optimal model with the correlation coefficient of prediction set (Rp) and root mean square error of prediction (RMSEP) of 0.936 and 0.828, respectively. The real-time evaluation of the spoilage was successfully obtained by loading an optimal monitoring and warning model into the microcontroller. An apple remote monitoring and early warning platform was built to visualize the apple warehouse's sensors data and spoilage level. The results demonstrated that the prototype based on characteristic gas sensor array could effectively monitor and warn apple spoilage.

Prediction of soil properties based on characteristic wavelengths with optimal spectral resolution by using Vis-NIR spectroscopy

Visible and near-infrared (Vis-NIR) spectroscopy technique has been recognized as a cost-effective, rapid, non-destructive alternative to traditional soil physicochemical analysis to estimate soil properties over the past few decades. Most efforts are devoted to the selection of characteristic wavelengths to eliminate the uninformative variables while ignoring the impact of the spectral resolution of these wavelengths on the prediction accuracy of soil properties. Therefore, the originality of this study is to identify the characteristic wavelengths with the optimal spectral resolution to achieve a better prediction performance. A 'two-step' wavelength selection method was proposed to select the characteristic wavelengths. Then, we simulated 1 nm-100 nm spectral resolution based on the spectral database measured by a portable ASD spectroradiometer and adopted the artificial bee colony (ABC) algorithm to further improve the prediction ability by configuring the most appropriate spectral resolution for each characteristic wavelength. The soil databases for this study consisted of 112 soil samples collected from Songnen Plain area in northeast China, and partial least squares regression (PLSR) was used to establish relations between pretreatment spectra and soil properties, including soil organic matter (SOM), available phosphorus (AP), and available potassium (AK). The independent validation results of this strategy effectively favored the prediction accuracy of SOM ( [Formula: see text] ), AP ( [Formula: see text] ), and AK ( [Formula: see text] ) compared with the PLSR models developed with full-spectra. In general, the method presented in this study suggested a framework for selecting characteristic wavelengths with optimal spectral solutions to predict SOM, AP, AK, and perhaps some other soil properties. The results of this paper also will provide guidance for the development of the low-cost specialized spectroscopic instruments for soil properties measurement.

Protected Geographical Indication Discrimination of Zhejiang and Non-Zhejiang Ophiopogonis japonicus by Near-Infrared (NIR) Spectroscopy Combined with Chemometrics: The Influence of Different Stoichiometric and Spectrogram Pretreatment Methods

This paper presents a method for the protected geographical indication discrimination of Ophiopogon japonicus from Zhejiang and elsewhere using near-infrared (NIR) spectroscopy combined with chemometrics. A total of 3657 Ophiopogon japonicus samples from five major production areas in China were analyzed by NIR spectroscopy, and divided into 2127 from Zhejiang and 1530 from other areas ('non-Zhejiang'). Principal component analysis (PCA) was selected to screen outliers and eliminate them. Monte Carlo cross validation (MCCV) was introduced to divide the training set and test set according to a ratio of 3:7. The raw spectra were preprocessed by nine single and partial combination methods such as the standard normal variable (SNV) and derivative, and then modeled by partial least squares regression (PLSR), a support vector machine (SVM), and soft independent modeling of class analogies (SIMCA). The effects of different pretreatment and chemometrics methods on the model are discussed. The results showed that the three pattern recognition methods were effective in geographical origin tracing, and selecting the appropriate preprocessing method could improve the traceability accuracy. The accuracy of PLSR after the standard normal variable was better, with R² reaching 0.9979, while that of the second derivative was the lowest with an R² of 0.9656. After the SNV pretreatment, the accuracy of the training set and test set of SVM reached the highest values, which were 99.73% and 98.40%, respectively. The accuracy of SIMCA pretreated with SNV and MSC was the highest for the origin traceability of Ophiopogon japonicus, which could reach 100%. The distance between the two classification models of SIMCA-SNV and SIMCA-MSC is greater than 3, indicating that the SIMCA model has good performance.

Non-Destructive Detection of Meat Quality Based on Multiple Spectral Dimension Reduction Methods by Near-Infrared Spectroscopy

The potential of four dimension reduction methods for near-infrared spectroscopy was investigated, in terms of predicting the protein, fat, and moisture contents in lamb meat. With visible/near-infrared spectroscopy at 400-1050 nm and 900-1700 nm, respectively, calibration models using partial least squares regression (PLSR) or multiple linear regression (MLR) between spectra and quality parameters were established and compared. The MLR prediction models for all three quality parameters based on the wavelengths selected by stepwise regression achieved the best results in the spectral region of 400-1050 nm. As for the spectral region of 900-1700 nm, the PLSR prediction model based on the raw spectra or high-correlation spectra achieved better results. The results of this study indicate that sampling interval shortening and of peak-to-trough jump features are worthy of further study, due to their great potential in explaining the quality parameters.

Discrimination of whole grape juice using fluorescence spectroscopy data with linear discriminant analysis coupled to genetic and ant colony optimisation algorithms

In this study, a new approach was developed for classifying grape juices produced in Brazil using unfolded excitation-emission matrix (EEM) fluorescence spectroscopy and chemometrics, with respect to the agricultural production system, namely the conventional or organic agricultural one. Linear discriminant analysis (LDA) coupled to ant colony optimisation (ACO) and the genetic algorithm (GA) were used to select a more effective subset of variables to discriminate grape juice samples. The best results demonstrated highly efficient classification of grape juice samples according to a conventional or organic production process with an accuracy rate of up to 97% for the models and 94% in the prediction of these classes for samples external to the model. The models showed high selectivity and sensitivity with a rate of up to 100% for the training and test datasets, in addition to determining the most significant variables that explain the separation of classes. The proposed method proves to be viable, as it is fast and requires minimal sample preparation, allowing quality control in the food industry.

Coal seam in-situ inorganic analysis based on least angle regression and competitive adaptive reweighted sampling algorithm by XRF-visNIR fusion

The fusion of X-ray fluorescence spectroscopy (XRF) and visible near infrared spectroscopy (visNIR) has been widely used in geological exploration. The outer product analysis (OPA) has a good effect in the fusion. The dimension of the spectral matrix obtained by OPA is large, and the Competitive Adaptive Reweighted Sampling (CARS) cannot cover the whole spectrum. As a result, the selected variables by the method are inconsistent each time. In this paper, a new feature variable screening method is proposed, which uses the Least Angle Regression (LAR) to select the high dimensional spectral matrix first, and then uses CARS to complete the secondary selection of the spectral matrix, forming the LAR-CARS algorithm. The purpose is to make the sampling method cover all the spectral data. XRF and visNIR tests were carried out on three cores in two boreholes, and a cross-validation set, validation set and a test set were established by combining the results of wavelength dispersion X-ray fluorescence spectrometer and ITRAX Core scanner in the laboratory. The quantitative model was established with the Extreme Gradient Boosting (XGBoost) and LAR-CARS was compared to these other algorithms (LAR, Successive Projections Algorithm, Monte Carlo uninformative variables elimination and CARS). The results showed that the RMSEP values of the models established by the LAR-CARS for six rock-forming elements (Si, Al, K, Ca, Fe, Ti) were relatively small, and the RPD ranges from 1.424 to 2.514. All these results show that the high-dimensional matrix formed by XRF and visNIR integration combined with LAR-CARS can be used for quantitative analysis of rock forming elements in in-situ coal seam cores, and the analysis results can be used as the basis for judging lithology. The research will provide necessary technical support for digital mine construction.

Freshness Identification of Oysters Based on Colorimetric Sensor Array Combined with Image Processing and Visible Near-Infrared Spectroscopy

Volatile organic compounds (VOCs) could be used as an indicator of the freshness of oysters. However, traditional characterization methods for VOCs have some disadvantages, such as having a high instrument cost, cumbersome pretreatment, and being time consuming. In this work, a fast and non-destructive method based on colorimetric sensor array (CSA) and visible near-infrared spectroscopy (VNIRS) was established to identify the freshness of oysters. Firstly, four color-sensitive dyes, which were sensitive to VOCs of oysters, were selected, and they were printed on a silica gel plate to obtain a CSA. Secondly, a charge coupled device (CCD) camera was used to obtain the “before” and “after” image of CSA. Thirdly, VNIS system obtained the reflected spectrum data of the CSA, which can not only obtain the color change information before and after the reaction of the CSA with the VOCs of oysters, but also reflect the changes in the internal structure of color-sensitive materials after the reaction of oysters’ VOCs. The pattern recognition results of VNIS data showed that the fresh oysters and stale oysters could be separated directly from the principal component analysis (PCA) score plot, and linear discriminant analysis (LDA) model based on variables selection methods could obtain a good performance for the freshness detection of oysters, and the recognition rate of the calibration set was 100%, while the recognition rate of the prediction set was 97.22%. The result demonstrated that the CSA, combined with VNIRS, showed great potential for VOCS measurement, and this research result provided a fast and nondestructive identification method for the freshness identification of oysters.

Improve the precision of platelet spectrum quantitative analysis based on "M+N" theory

Platelets have the functions of promoting blood coagulation and accelerating hemostasis, playing an important role in human body. It is of great medical significance to realize clinical rapid micro-detection of platelets by spectral analysis, which is the development direction of clinical detection in the future. However, due to the problem of unobvious characteristic of platelet absorption spectrum, the results of modeling and analysis cannot meet the clinical accuracy requirements. In order to improve the analysis accuracy, based on the "M+N" theory, this paper comprehensively considers the influence of the concentrations of measured component platelet and non-measured component hemoglobin on modeling analysis, and uses the method of selecting training set based on the concentration distribution of two components. At the same time, considering the characteristic of the linear model, the samples at both ends of the concentration of two components are selected as the training set, and the cubic term fitting method is combined to model and predict the concentration of platelet. The following experiments were designed: the training sets were selected by four different methods and used for modeling to predict the platelet concentration, and compared the modeling results of different methods. Through the modeling and prediction of 222 samples, the result showed that the method of selecting the training set with the concentration distribution of two components could effectively improve the prediction accuracy of the established model, and got a better model with better performance, the correlation coefficient Rc reached 0.63, which was 24.98% higher than the result of full modeling for all samples, and RMSE decreased by 10.02%. Considering the influence of non-measured components in modeling is of great significance to improve the prediction accuracy of measured components, and selecting samples from both ends of the concentration values of two components as the training set can further improve the performance and accuracy of the model.

Label-free Au NRs-based SERS coupled with chemometrics for rapid quantitative detection of thiabendazole residues in citrus

Citrus is a highly consumed fruit worldwide. However, the excessive use of thiabendazole (TBZ) pesticides during citrus cultivation poses a health risk to people. Hence, a rapid and quantitative method has been established for TBZ determination in citrus by coupling gold nanorods (Au NRs) based surface enhanced Raman scattering (SERS) coupling chemometrics. The results show that support vector machine (SVM) can distinguish TBZ residues of different orders of magnitude with 99.1667% accuracy and that genetic algorithm-partial least squares (GA-PLS) had the best performance in the quantitative prediction of TBZ residues (Rp² = 0.9737, RMSEP = 0.1179 and RPD = 5.85) in citrus. The limit of detection (LOD) was 0.33 μg mL^-1. Furthermore, the proposed method was validated by a standard HPLC method using t-test with no significant difference. Therefore, the proposed Au NRs-based SERS technique can be used for the rapid quantitative analysis of TBZ residues in citrus.

Classification of soybean frogeye leaf spot disease using leaf hyperspectral reflectance

In this study, the feasibility of classifying soybean frogeye leaf spot (FLS) is investigated. Leaf images and hyperspectral reflectance data of healthy and FLS diseased soybean leaves were acquired. First, image processing was used to classify FLS to create a reference for subsequent analysis of hyperspectral data. Then, dimensionality reduction methods of hyperspectral data were used to obtain the relevant information pertaining to FLS. Three single methods, namely spectral index (SI), principal component analysis (PCA), and competitive adaptive reweighted sampling (CARS), along with a PCA and SI combined method, were included. PCA was used to select the effective principal components (PCs), and evaluate SIs. Characteristic wavelengths (CWs) were selected using CARS. Finally, the full wavelengths, CWs, effective PCs, SIs, and significant SIs were divided into 14 datasets (DS1-DS14) and used as inputs to build the classification models. Models' performances were evaluated based on the classification accuracy for both the overall and individual classes. Our results suggest that the FLS comprised of five classes based on the proportion of total leaf surface covered with FLS. In the PCA and SI combination model, 5 PCs and 20 SIs with higher weight coefficient of each PC were extracted. For hyperspectral data, 20 CWs and 26 effective PCs were also selected. Out of the 14 datasets, the model input variables provided by five datasets (DS2, DS3, DS4, DS10, and DS11) were more superior than those of full wavelengths (DS1) both in support vector machine (SVM) and least squares support vector machine (LS-SVM) classifiers. The models developed using these five datasets achieved overall accuracies ranging from 91.8% to 94.5% in SVM, and 94.5% to 97.3% in LS-SVM. In addition, they improved the classification accuracies by 0.9% to 3.6% (SVM) and 0.9% to 3.7% (LS-SVM).

Nondestructive Analysis of Internal Quality in Pears with a Self-Made Near-Infrared Spectrum Detector Combined with Multivariate Data Processing

The consumption of pears has increased, thanks not only to their delicious and juicy flavor, but also their rich nutritional value. Traditional methods of detecting internal qualities (e.g., soluble solid content (SSC), titratable acidity (TA), and taste index (TI)) of pears are reliable, but they are destructive, time-consuming, and polluting. It is necessary to detect internal qualities of pears rapidly and nondestructively by using near-infrared (NIR) spectroscopy. In this study, we used a self-made NIR spectrum detector with an improved variable selection algorithm, named the variable stability and cluster analysis algorithm (VSCAA), to establish a partial least squares regression (PLSR) model to detect SSC content in snow pears. VSCAA is a variable selection method based on the combination of variable stability and cluster analysis to select the infrared spectrum variables. To reflect the advantages of VSCAA, we compared the classical variable selection methods (synergy interval partial least squares (SiPLS), genetic algorithm (GA), successive projections algorithm (SPA), and bootstrapping soft shrinkage (BOSS)) to extract useful wavelengths. The PLSR model, based on the useful variables selected by SiPLS-VSCAA, was optimal for measuring SSC in pears, and the correlation coefficient of calibration (Rc), root mean square error of cross validation (RMSECV), correlation coefficient of prediction (Rp), root mean square error of prediction (RMSEP), and residual predictive deviation (RPD) were 0.942, 0.198%, 0.936, 0.222%, and 2.857, respectively. Then, we applied these variable selection methods to select the characteristic wavelengths for measuring the TA content and TI value in snow pears. The prediction PLSR models, based on the variables selected by GA-BOSS to measure TA and that by GA-VSCAA to detect TI, were the best models, and the Rc, RMSECV, Rp and RPD were 0.931, 0.124%, 0.912, 0.151%, and 2.434 and 0.968, 0.080%, 0.968, 0.089%, and 3.775, respectively. The results showed that the self-made NIR-spectrum detector based on a portable NIR spectrometer with multivariate data processing was a good tool for rapid and nondestructive analysis of internal quality in pears.

Data Fusion of XRF and Vis-NIR Using Outer Product Analysis, Granger–Ramanathan, and Least Squares for Prediction of Key Soil Attributes

Visible-near-infrared (vis-NIR) and X-ray fluorescence (XRF) are key technologies becoming pervasive in proximal soil sensing (PSS), whose fusion holds promising potential to improve the estimation accuracy of soil attributes. In this paper, we examine different data fusion methods for the prediction of key soil fertility attributes including pH, organic carbon (OC), magnesium (Mg), and calcium (Ca). To this end, the vis-NIR and XRF spectra of 267 soil samples were collected from nine fields in Belgium, from which the soil samples of six fields were used for calibration of the single-sensor and data fusion models while the validation was performed on the remaining three fields. The first fusion method was the outer product analysis (OPA), for which the outer product (OP) of the two spectra is computed, flattened, and then subjected to partial least squares (PLS) regression model. Two versions of OPA were evaluated: (i) OPA-FS in which the full spectra were used as input; and (ii) OPA-SS in which selected spectral ranges were used as input. In addition, we examined the potential of least squares (LS) and Granger–Ramanathan (GR) analyses for the fusion of the predictions provided by the single-sensor PLS models. Results demonstrate that the prediction performance of the single-sensor PLS models is improved by GR in addition to the LS fusion method for all soil attributes since it accounts for residuals. Resorting to LS, the largest improvements compared to the single-sensor models were obtained, respectively, for Mg (residual prediction deviation (RPD) = 4.08, coefficient of determination (R2) = 0.94, ratio of performance of inter-quantile (RPIQ) = 1.64, root mean square error (RMSE) = 4.57 mg/100 g), OC (RPD = 1.79, R2 = 0.69, RPIQ = 2.82, RMSE = 0.16%), pH (RPD = 1.61, R2 = 0.61, RPIQ = 3.06, RMSE = 0.29), and Ca (RPD = 3.33, R2 = 0.91, RPIQ = 1, RMSE = 207.48 mg/100 g). OPA-FS and OPA-SS outperformed the individual, GR, and LS models for pH only, while OPA-FS was effective in improving the individual sensor models for Mg as well. The results of this study suggest LS as a robust fusion method in improving the prediction accuracy for all the studied soil attributes.

Rapid detection of cellulose and hemicellulose contents of corn stover based on near-infrared spectroscopy combined with chemometrics

The feasibility of near-infrared spectroscopy (NIRS) combined with chemometrics for the rapid detection of the cellulose and hemicellulose contents in corn stover is discussed. Competitive adaptive reweighted sampling (CARS) and genetic simulated annealing algorithm (GSA) were combined (CARS-GSA) to select the characteristic wavelengths of cellulose and hemicellulose and to reduce the dimensionality and multicollinearity of the NIRS data. The whole spectra contained 1845 wavelength variables. After CARS-GSA optimization, the number of characteristic wavelengths of cellulose (hemicellulose) was reduced to 152 (260), accounting for 8.24% (14.09%) of all wavelengths. The coefficients of determination of the regression models for predicting the cellulose and hemicellulose contents were 0.968 and 0.996, the root mean square errors of prediction (RMSEPs) were 0.683 and 0.648, and the residual predictive deviations (RPDs) were 5.213 and 16.499, respectively. The RMSEP of the cellulose and hemicellulose regression models was 0.152 and 0.190 lower for CARS-GSA than for the full-spectrum, and the RPD was increased by 0.949 and 3.47, respectively. The results showed that the CARS-GSA model substantially reduced the number of characteristic wavelengths and significantly improved the predictive ability of the regression model.

Wine Authenticity and Traceability with the Use of FT-IR

Fourier transform infrared spectroscopy (FT-IR) has gained popularity in the wine sector due to its simplicity and ability to provide a wine’s fingerprint. For this reason, it is often used for authentication and traceability purposes with more than satisfactory results. In this review, an outline of the reasons why authenticity and traceability are important to the wine sector is given, along with a brief overview of the analytical methods used for their attainment; statistical issues and compounds, on which authentication usually is based, are discussed. Moreover, insight on the mode of action of FT-IR is given, along with successful examples from its use in different areas of interest for classification. Finally, prospects and challenges for suggested future research are given. For more accurate and effective analyses, the construction of a large database consisting of wines from different regions, varieties and winemaking protocols is suggested.

Improving the analysis accuracy of components in blood by SSP-MCSD and multi-mode spectral data fusion

In recent years, spectral quantitative analysis for blood components has been a research hotspot in biomedical engineering. But researches have been limited to the application of high-sensitivity spectroscopy instruments and the complexity of blood components-the overlapping of absorption curves for many components is severe. This has led to the difficulty in achieving satisfactory results when using spectroscopy to quantify components in blood. In order to enhance the model robustness and improve the model performance, this paper proposed a sample set partitioning strategy based on multi-component spatial distance (SSP-MCSD). Different from the other sample set partitioning strategies, which only consider the uniformity of the concentration distribution of the target component, this strategy also concerns to the concentration distribution of non-target components. The concentration of the target component and non-target components are used to construct a multi-dimensional space, and the Euclidean Distance of sample points in this space is used as the criterion to partition the sample set. At the same time, the spectra collected in multi-modes are fused for increasing the amount of information. So as to enhance the model robustness and to improve the analysis accuracy of the target components. In order to verify the effectiveness of this strategy, the serum of 101 volunteers was analyzed. Taking total protein in serum as the non-target component, the regression model for bilirubin concentration was established by transmission spectra, fluorescence spectra, and the joint spectra after fusion of the above two spectra, respectively. The experimental results showed that the prediction accuracy of the model established by SSP-MCSD combined with multi-mode spectral fusion is obviously higher than that of other methods. It can effectively improve the analysis accuracy of blood components.

Accuracy improvement of quantitative analysis in VIS-NIR spectroscopy using the GKF-WTEF algorithm

The extraction of effective information in visible-near-infrared (VIS-NIR) spectroscopy is crucial and difficult for spectral analysis. In this research, an algorithm of wavelet feature extraction based on the Gaussian kernel function (GKF-WTEF) was developed to suppress the influence of external interference on VIS-NIR spectroscopy and improve the accuracy of quantitative analysis. This algorithm takes the root-mean-square error of the prediction set (RMSEP) of the model, which is established by partial least-squares regression, as the optimization criteria. First, the optimal type of wavelet function, the decomposition level, and the Gauss kernel function central frequency band are determined according to the RMSEP. Second, the Gauss kernel function bandwidth is determined by Newton's method. Then, the Hadamard product of the Gaussian kernel function and the wavelet coefficient is obtained. Finally, the wavelet coefficients after the Hadamard product can be reconstructed to obtain the spectral data after feature extraction. In order to verify the effectiveness of this algorithm, the difference in the optical parameters of the polyvinyl chloride material container was used as an external interference source. And the spectrum of Intra-lipid and India-ink mixed solution with different concentrations was collected therein. The volume fraction of India-ink in complex mixed solution was quantitatively analyzed by using the RMSEP and the average relative error of the prediction set as the evaluation criteria. The research results demonstrated that the Gaussian-wavelet transform feature extraction algorithm is an effective pretreatment method, it can satisfactorily suppress the influence of external interference on the spectrum, and it can improve the analytical accuracy of VIS-NIR spectroscopy.

Near infrared system coupled chemometric algorithms for the variable selection and prediction of baicalin in three different processes

Characteristic variables are essential and necessary basis in model construction, and are related to the prediction result closely in near infrared spectroscopy (NIRS) analysis. However, the same compound usually has different characteristic variables for different analysis and it would be lower correlation between variables and structure in many researches. So, the accuracy and reliability are expected to improve by exploring characteristic variables in different spectrum analysis. In this study, competitive adaptive weighted resampling method (CARS) was applied to select characteristic variables related to baicalin from NIRS analysis data, which were applied to analysis of baicalin in three different processes including the herb, extraction process and concentration process of Scutellaria baicalensis. After application of CARS method, 70, 50 and 50 variables were selected respectively from three processes above. The selected variables were firstly analyzed by statistical methods that they were found to be consistent and correlated among three different processes after one-way analysis of variance test and Kendall's W. Partial least-squares (PLS) regression and extreme learning machine (ELM) models were constructed based on optimized data. Models after variable selection were less complicated and had better prediction results than global models. After comparison, CARS-PLS was suitable for the prediction of extraction process, while for the concentration process and herb, CARS-ELM performed better. The Rc value of the herb, extraction and concentration model were 0.9469, 0.9841 and 0.9675, respectively. The RSEP values were 4.54%, 6.96% and 8.37%, respectively. The results help to frame a theoretical basis for characteristic variables of baicalin.

Prediction Models to Control Aging Time in Red Wine

A combination of physical-chemical analysis has been used to monitor the aging of red wines from D.O. Toro (Spain). The changes in the chemical composition of wines that occur over the aging time can be used to distinguish between wine samples collected after one, four, seven and ten months of aging. Different computational models were used to develop a good authenticity tool to certify wines. In this research, different models have been developed: Artificial Neural Network models (ANNs), Support Vector Machine (SVM) and Random Forest (RF) models. The results obtained for the ANN model developed with sigmoidal function in the output neuron and the RF model permit us to determine the aging time, with an average absolute percentage deviation below 1%, so it can be concluded that these two models have demonstrated their capacity to predict the age of wine.