Data-driven Soft Independent Modeling Of Class Analogy Research Articles

Evaluation of adulteration in soy-based beverages by water addition using chemometrics applied to ATR-FTIR spectroscopy

The adulteration of soy-based beverages (SBBs) by adding water to increase profitability is a fraudulent practice that requires urgent solutions to ensure product integrity and consumer trust. Therefore, the use of infrared spectroscopy (ATR-FTIR) associated with chemometrics methods can be a quick and advantageous alternative to this problem. In this study, the one-class and multiclass methods applied to ATR-FTIR data to classify a set of 80 SBBs samples were used. The unequal dispersed classes (UNEQ), soft independent modeling of class analogy (SIMCA), data driven SIMCA (DD-SIMCA), and one-class random forest (OC-RF) methods were used for one-class modeling. Models were constructed using the non-adulterated samples as target class (TA) and the adulterated samples as non-target class (NT). The k-nearest neighbors (k-NN), partial least squares discriminant analysis (PLS-DA), dual class random forest (DC-RF), and dual class random forest with Monte Carlo sampling (DC-RF-MC) methods were used for multiclass modeling. For k-NN and PLS-DA, samples were organized into four classes (non-adulterated samples, adulterated with 5% v∙v-1, 10% v∙v-1, and 20% v∙v-1 of water). DC-RF models used the same class settings as one-class models. DD-SIMCA, PLS-DA, and DC-RF-MC showed accuracy of 100%. The results show the feasibility of ATR-FTIR and chemometrics models to identify adulterations by adding water.

Geographic authentication of argentinian teas by combining one-class models and discriminant methods for modeling near infrared spectra

In this study, 110 tea samples from South American countries (Argentina, Brazil, and Paraguay) and Asian countries (India and China) were analyzed using near-infrared spectroscopy (NIRS) together with a two-step chemometric authentication strategy (class modeling techniques and discriminant analysis) to authenticate commercial teas from Argentina. In the first step, one-class models were built and validated to authenticate South American teas using preprocessed NIRS data. For this purpose, data-driven soft independent modeling of class analogy (DD-SIMCA) and one-class partial least squares (OC-PLS) were used. The DD-SIMCA model gave the best results, with a sensitivity of 93.10%, specificity of 100%, and efficiency of 95.00%. In the second step, a support vector machine (SVM) was used to build and validate a multiclass model to discriminate between tea samples from Argentina and neighboring countries of South America. The best model was the combination of nine variables selected by the fast correlation-based filter (FCBF) method, with an accuracy of 98.30%. Therefore, we conclude that the combination of NIRS and two-step chemometric tools can be used to authenticate the geographical origin of samples with high inter-class similarity.

Data-Driven Version of Multiway Soft Independent Modeling of Class Analogy (N-Way DD-SIMCA): Theory and Application.

One-class classification (OCC) is discussed in the framework of the measurement and processing of multiway data. Data-driven soft independent modeling of class analogy (DD-SIMCA) is applied in the following formats: (1) multiblock and (2) Tucker 3 N-way SIMCA, which are shown to be useful tools for solving classification tasks. A new decision rule for N-way DD-SIMCA is adopted based on the conventional two-way DD-SIMCA model. Multiblock SIMCA is shown to be useful for variable selection, and Tucker 3 SIMCA to select the optimal model complexity when applying multiway data decomposition and to assess the role of individual samples in the classification model. Both approaches, together with the two-way DD-SIMCA version applied to the unfolded data, are compared regarding the analysis of an experimental data set including genuine and adulterated blueberry extract samples. The latter were employed to produce matrix spectral-time data matrices per sample within a flow injection system, taking advantage of the spectral changes in the sample constituents as a function of the pH of the carrier phase. The need to employ the Tucker 3 model instead of a trilinear decomposition is supported by a discussion on the lack of the trilinearity property of the studied data.

Authentication of Specialty Coffees from the Fluminense Northwest and Caparaó Regions (Brazil) Using UV-Vis Spectroscopy and Synthetic Samples Partial Least Square Discriminant Analysis (SS-PLS-DA)

Caparaó and the Fluminense northwest regions are nationally recognized by the important contribution on coffee production and exportation. Adulterations involving specialty coffees result in a decrease in the quality of the final product. However, obtaining many different samples from the same region is unfeasible in some cases, needing strategies to work with a limited number of samples for pattern recognition. Thus, this work is the first to use the construction of synthetic samples (SS) for analysis of coffees, and its objective is to identify adulterations in specialty coffees with bark, straw and low-quality beans, using UV-Vis spectroscopy, associated with chemometric methods. The synthetic samples partial least square discriminant analysis (SS-PLS-DA) showed better specificity, sensitivity and reliability rates than the Hard PLS-DA models. One-class methods (soft independent modeling of class analogy (SIMCA) and data driven soft independent modeling of class analogy (DD-SIMCA)) showed low specificity and reliability. The discriminant methods together with the synthetic samples proved to be adequate to identify adulterations in specialty coffees.

Limit of detection in qualitative analysis: Classification Analytical Signal approach

This paper presents a new methodology for the estimation of the limit of detection and other related parameters in qualitative analysis. Estimation of the detection limit is important in special cases when a small amount of an illegal ingredient can significantly affect the quality of the product or the presence of even a small amount of an extraneous analyte is significant. The approach is based on a recently introduced concept of an analytical signal in qualitative analysis, called the Classification Analytical Signal (CAS). A detailed algorithm to calculate the detection capability for a one-class classification method Data-Driven Soft Independent Modeling of Class Analogy (DD-SIMCA) is described and illustrated with two real-world examples. Both examples consider cases where the detection of low analyte concentrations is important and corresponding samples should be considered as non-compliant. The first case presents adulteration of wheat flour with talcum powder, where the samples were specially prepared with various concentrations of the analyte. The second case uses a publicly available dataset devoted to the estimation of tetracycline in blood serum. The latter case is used to compare the limits of detection estimated for multivariate calibration with those presented for qualitative analysis.

Fingerprinting based on spectral reflectance and chemometrics – An analytical approach aimed at combating the illegal trade of stingray meat in the Amazon

The survival of Amazon stingrays is threatened due to excessive fishing and habitat degradation. To address this issue, this study developed a groundbreaking method to authenticate and differentiate Amazon stingray meats using a portable spectrophotometer and chemometrics. Samples were collected from various species, including an endangered one with a commercialization ban and no population reduction records. Principal Component Analysis (PCA), identified natural groupings based on the meat's commercial origin, while Partial Least Squares-Discriminant Analysis (PLS-DA), accurately discriminated the commercial and geographic origins with 100 % accuracy. Moreover, Data-Driven Soft Independent Modeling of Class Analogy (DD-SIMCA), effectively distinguished Amazon stingray meat from other marketable species. This approach offers a rapid, precise, and non-destructive means for monitoring and controlling the illegal trade of these species, thereby supporting decision-making in the field and promoting the conservation and sustainability of freshwater stingrays in the Amazon region.

Rapid authentication of intact saffron stigma thorough the package using Vis-SWNIR hyperspectral imaging coupled with chemometrics

This study presents a rapid method for detecting and authenticating intact saffron stigma through its packaging using visible-short wavelength near-infrared (Vis-SWNIR) hyperspectral imaging (HSI) in conjunction with chemometrics. A dataset containing 38 authentic saffron samples was utilized, and HSI images were captured within the spectral range of 400–950 nm. Multivariate curve resolution-alternating least squares (MCR-ALS) was used to obtain pure spatial and spectral profiles of the components. The resolved spatial profiles were then subjected to principal component analysis (PCA) to identify potential patterns in the authentic samples. Binary mixtures of pooled saffron samples and adulterants, including safflower, calendula, and saffron style, were prepared at varying levels. Data-driven soft independent modeling of class analogy (DD-SIMCA) was employed to establish a boundary between authentic and adulterated samples, resulting in robust classification performance with a sensitivity of 95 % and specificity of 100 %. Partial least squares-discriminant analysis (PLS-DA) was used for binary and multi-class classification of the MCR-ALS resolved spatial profiles. The performance of PLS-DA was evaluated using sensitivity, specificity, and accuracy measures, which exceeded 83.4 % for the calibration, cross-validation, and prediction sets across three different saffron packaging methods. These findings provide further evidence for the validity and efficacy of the proposed method.

Combining NIR spectroscopy with DD‐SIMCA for authentication and iSPA‐PLS‐DA for discrimination of ethyl route and oil feedstocks of biodiesels in biodiesel/diesel blends

AbstractThe main biofuels produced on an industrial large scale are biodiesel and ethanol, which are the most economically viable and widely implemented solutions to replace conventional fossil fuels from a greener and more sustainable perspective. In such a scenario, there is an opportunity to produce fully renewable biodiesel using ethanol instead of methanol, which is mainly derived from fossil resources. In this paper, near‐infrared (NIR) spectroscopy was used to discriminate biodiesel/diesel (B7) blends regarding the synthesis route and oil feedstock of biodiesels simultaneously. Data‐Driven Soft Independent Modeling of Class Analogy (DD‐SIMCA) authenticated correctly all ethyl B7 (target) samples into the acceptance area, while rejected all non‐target samples, implying in an efficiency of 100%. Additionally, Partial Least Squares‐Discriminant Analysis coupled with interval selection by the Successive Projections Algorithm (iSPA‐PLS‐DA) discriminated all ethyl B7 samples correctly, considering cottonseed, sunflower, and soybean as oil feedstocks. Moreover, only one ethyl cottonseed B7 sample was incorrectly discriminated when methyl B7 samples from the same three oil feedstocks were included in the model. As advantages, the proposed analytical methodology contributes to the United Nations' Sustainable Development Goal (SDG) #7 (affordable and clean energy) as well as aligns with the principles of Green Analytical Chemistry.

Non-destructive analytical strategies based on NIR spectroscopy for monitoring the quality of biodiesel/diesel blends in terms of the authentication of second-generation biodiesel feedstock using DD-SIMCA and quantification of the biodiesel content using iSPA-PLS

Considering relevant environmental, economic, and social factors, the discussion on the use of second-generation biofuels has gained prominence in recent years, especially supported by the Sustainable Development Goals (SDGs) and the Paris COP21 climate change agreement. One of the policies implemented by different countries has been the mandatory blend of biodiesel in commercial petrodiesel. Thus, this work proposes the use of NIR spectroscopy to monitor the quality of biodiesel/diesel blends. The analytical strategies involve the authentication of second-generation biodiesel feedstock using Data-Driven Soft Independent Modeling of Class Analogy (DD-SIMCA) and the quantification of biodiesel content regardless of the biodiesel feedstock using the Successive Projections Algorithm for interval selection in Partial Least Squares (iSPA-PLS). To avoid dilution of the samples with organic solvents, two spectral regions (881–1651 and 1911–2203 nm) were investigated. As a result, the spectral range of 1911–2203 nm achieved the best results for both the qualitative and quantitative analysis. Accordingly, the authentication efficiencies for DD-SIMCA were 97.6% and 99.2% for jatropha and beef tallow biodiesel/diesel blends using multiplicative scattering correction (MSC) and Savitzky-Golay smoothing (with a second-order polynomial and a 15-point window) coupled with the first derivative (SGD) as pre-processing algorithms, respectively. Furthermore, a relative error of prediction of only 2.85% was obtained for the biodiesel content quantification using SGD/iSPA-PLS. Hence, the proposed analytical methodology provides valuable support to governments and regulatory bodies in overseeing the quality of this product, ensuring that consumers are safeguarded against economic repercussions resulting from improper or fraudulent blending. Moreover, it aligns with the principles of Green Analytical Chemistry and contributes to the SDG 7 (affordable and clean energy) and SDG 13 (climate action).

An innovative screening approach for orange juice authentication using dual portable/handheld NIR spectrometers and chemometrics

In this study, we explored the feasibility of utilizing two portable/handheld near-infrared (NIR) spectrometers in combination with class modeling techniques, namely data-driven soft independent modeling of class analogy (DD-SIMCA) and soft-partial least squares-discriminant analysis (soft-PLS-DA), as well as discrimination strategies of ensemble learning and hard-PLS-DA, for developing a screening method to authenticate orange juice samples and detect the presence of adulterants (the Brix to citric acid ratio in pulp-wash). The results obtained with both NIR spectrometers (Tellspec, 900-1700 nm and Neospectra, 1350-2550 nm) coupled with DD-SIMCA exhibited perfect sensitivity and specificity of 100% in both calibration and prediction sets. Additionally, when combined with the NIR Tellspec spectrometer, ensemble learning methods including gradient boosting tree (GBT) and adaptive boosting (Adaboost) demonstrated exceptional predictive capabilities, achieving the prediction of 'Brix to citric acid' ratio in pulp-wash with high sensitivity and specificity. These results surpassed the performance of hard-PLS-DA as commonly used technique in food authentication.

Geographical origin identification of camellia oil based on fatty acid profiles combined with one-class classification

Geographical Indication (GI) agricultural products possess specific geographical origins and high qualities, which require an effective geographical origin traceability method for the important protective trademarks. In this study, authentication models for Changshan camellia oil were developed by fatty acid profiles and one-class classification methods including data-driven soft independent modeling of class analogy (DD-SIMCA) and one-class partial least squares (OCPLS), and compared with traditional two-class classification models. The results indicated that the prediction errors of three two-class classification models were 63.8%, 12.1%, and 65.2% for the samples out of targeted geographical origins, respectively. By contrast, the one-class classification models could completely differentiate Changshan from non-Changshan camellia oils, even from the adjacent counties. Moreover, compared with traditional indicators of mineral elements, the model built by fatty acid profiles possessed higher sensitivity and specificity. It also offered a reference strategy for the geographical origin identification of other high-value oils or foods.

Technical exploration of the origins, storage periods and species identification of Boletus bainiugan

To prevent fraud in Boletus bainiugan commodities, this study provides the market with two fast and stable identification models for accurate identification of Boletus bainiugan origins, storage periods and species. Partial least squares discrimination analysis (PLS-DA), Support vector machine (SVM), Residual convolutional neural network (ResNet) and Data-driven soft independent modeling of class analogy (DD-SIMCA) models were built by combining with Fourier transform near-infrared spectroscopy (FT-NIR). The results show that the ResNet model is significant in solving the Boletus bainiugan origin identification problems. The ResNet model had the best performance and highest accuracy compared to the PLS-DA and SVM models. The DD-SIMCA model was the preferred method for the one-class classification problem, achieving an accuracy of over 96% for the Boletus bainiugan storage period and species identification. Non-target class classification accuracy reached 100%. In summary, FT-NIR combined with ResNet and DD-SIMCA models were able to solve the related identification problems of Boletus bainiugan with more satisfactory results.

Intrinsic SERS Fingerprints of Aptamer-Peptide Conjugates for Direct High-Specific Profiling Abnormal Protein Levels in Cancer Patients.

Surface-enhanced Raman spectroscopy (SERS) with ultrasensitive vibrational fingerprints enables quick identification and trace detection of various kinds of molecules. But proteins usually have low Raman cross sections and are difficult to generate recognizable signals in direct SERS detection. Recently, nucleic acid-peptide conjugates are emerging with great potential in structuring, assembling, catalyzing, sensing, etc., and the coupling of aptamers further enables superior biological recognition and programmability. Here, we develop the aptamer-peptide conjugates as a new kind of SERS probe for direct high-specific profiling abnormal protein levels in cancer patients. The aptamer conjugated with glutathione (GSH) functions as both the recognition element and the SERS reporters that can simultaneously generate SERS fingerprints of both peptides and nucleic acids. This kind of biocompatible probe appears to have excellent performance in high-salt environments and realizes rapid, simple, and multisignal detection of thrombin (TB). Data-driven soft independent modeling of class analogy (DD-SIMCA) is used to distinguish SERS profiles of actual blood samples and realize the identification and classification of cancer patients. Furthermore, the effect of low-temperature storage time on blood samples is analyzed by tracking the changes of SERS profiles; the results hint that plasma samples stored under 4 °C for more than 2 days could generate false negative results due to TB hydrolysis, which has important implications for clinical sample analysis. This kind of nucleic acid-peptide conjugate provides new ideas for SERS sensing strategy in the future.

Using ATR-FTIR spectroscopy and DD-SIMCA for ecstasy profiling

Ecstasy is a recreational drug known to have serious adverse effects, necessitating the development of new techniques for its rapid and affordable identification in harm reduction policies and forensic laboratories. In this study, ATR-FTIR spectroscopy and DD-SIMCA (Data Driven Soft Independent Modeling of Class Analogy) were employed to determine the profile of ecstasy tablets and their adulterants. ATR-FTIR spectroscopy, which is a fast and non-destructive method that provides structural information, and DD-SIMCA, a supervised technique that distinguishes between different groups, were used. The Hierarchical Cluster Analysis (HCA) revealed the formation of two primary clusters: one comprising mainly MDA samples with some MDMA samples, and the second including cocaine, methamphetamine, and caffeine samples, as well as a mixture of MDA samples with methamphetamine and caffeine. Principal Component Analysis (PCA) was performed, and the first five principal components (PC) explained 86.25% of the total data variance. In the PC1xPC2 scores plot, the distribution of samples containing MDs (MDA and MDMA) was observed. Further analysis using PCA revealed the grouping of cocaine samples into two separate groups based on variations in concentration. The DD-SIMCA model demonstrated high sensitivity values in accurately identifying target samples (MDA) and relatively high specificity values in both the training and test sets. The study highlights the effectiveness of ATR-FTIR spectroscopy, PCA, and DD-SIMCA for the precise classification of ecstasy and its adulterants and suggests their potential in drug identification and analysis.

UV-Vis spectroscopy and one-class modeling for the authentication of the geographical origin of green coffee beans from Cerrado Mineiro, Brazil

Green coffee beans produced in the Cerrado Mineiro region have a Protected Designation of Origin (PDO) certificate. This guarantees that its coffees are among Brazil's most appreciated and valued. This study aimed to develop an authentication method to characterize and differentiate coffee beans from this region. Ultraviolet-visible spectra of 130 green coffee bean samples were used for one-class modeling with SIMCA (soft independent modeling of class analogy), DD-SIMCA (data-driven SIMCA), and OCPLS (one-class partial least squares). Ordered predictors selection (OPS) variable selection was applied to simplify multivariate models and improve their performance. The interpretation of developed models was performed by evaluating the parameter modeling power for variables of SIMCA models. It was possible to assign the most predictive variables to trigonelline and chlorogenic acids as responsible for discriminating coffee beans from Cerrado Mineiro from coffee beans of their three main competitor regions: Caparaó, Mogiana, and Sul de Minas. Finally, methodologies as developed in this work, based on one-class modeling methods, can be proposed to authenticate other coffee-producing regions, food matrices, and agricultural products.

Authentication and detection of common adulterants in clove buds (Syzygium aromaticum L.) powders and oils using near IR combined to multivariate analysis

This study investigates the use of NIR diffuse reflectance spectroscopy with multivariate analysis for quality control and authentication of clove buds powders as well as their oils. Unsupervised and supervised chemometric analysis techniques, including principal component analysis (PCA), data driven soft independent modeling of class analogy (DD-SIMCA) were implemented to authenticate clove and its oil and distinguish them from their adulterants. The SIMCA model showed 100% sensitivity and specificity values in detecting adulterants in clove powder samples and clove oil samples. This showcases the accuracy and reliability of the DD-SIMCA approach in accurately distinguishing between different classes.The models were validated using test samples, and the absence of noise modeling was confirmed through permutation. PLS regression analysis was utilized to measure levels of adulterants in clove powder and clove oil samples. The models produced good results, with RMSEC values ranging from 0.68 to 1.5% for clove oil and from 0.96 to 1.27% for clove powder. External validation was performed and the limits of quantitation ranged from 1.8% to 4.9% for clove powder and clove oil. The models can detect sample adulteration and ensure authenticity. The results showed that the suggested method and models can be used to detect sample adulteration, ensure authenticity, and have high sample throughput. The method had several advantages, including simplicity, speed of analysis, with no sample preparation needed.

Multi-block DD-SIMCA as a high-level data fusion tool

Multi-block classification method based on the Data Driven Soft Independent Modeling of Class Analogy (DD-SIMCA) is presented. A high-level data fusion approach is used for the joint analysis of data collected with the help of different analytical instruments. The proposed fusion technique is very simple and straightforward. It uses a Cumulative Analytical Signal which is a combination of outcomes of the individual classification models. Any number of blocks can be combined. Although the high-level fusion eventually leads to a rather complex model, the analysis of partial distances makes it possible to establish a meaningful relationship between the classification results and the influence of individual samples and specific tools.Two real world examples are used to demonstrate the applicability of the multi-block algorithm and the consistency of the multi-block method with its predecessor, a conventional DD-SIMCA.

Classification of instant coffees based on caffeine content and roasting degree using NIR spectrometry and multivariate analysis

Instant coffee is a beverage obtained from the dehydration of roasted coffee extract. Its quality depends on several factors, such as the type of bean and grinding, drying decaffeination, and roasting degree processes. In this work, NIR spectrometry and multivariate classification tools were combined to develop, for the first time, methods for authenticating and checking the compliance of decaffeinated instant coffees in order to ensure consumer food safety, as well as discriminating between regular instant coffees according to the roasting degree (traditional and extra-strong) for monitoring the production chain of regular instant coffees. First, one-class methods were built and validated to authenticate decaffeinated instant coffees and to in order to alert caffeine-sensitive consumers to potential risks. For this purpose, two near-infrared (NIR) instruments (benchtop and portable) and Data-Driven Soft Independent Modeling of Class Analogy (DD-SIMCA) and One-Class Partial Least Squares (OC-PLS) were used. After Partial Least Squares – Discriminant Analysis (PLS-DA), Successive Projections Algorithm for Interval Selection in Partial Least Squares – Discriminant Analysis (iSPA-PLS-DA), and the same NIR instruments were used to build and validate multiclass methods to discriminate regular instant coffees according to the roasting degree (traditional and extra-strong). The result for the classification of decaffeinated and roasting degree instant coffees was 100% and 98% correct classification rate using DD-SIMCA and iSPA-PLS-DA, respectively. Therefore, the combination of NIR spectroscopy and multivariate classification techniques proved valuable and fast tools for checking the quality of instant coffees.

DD-ComDim: A data-driven multiblock approach for one-class classifiers

An attractive alternative to solve authentication challenges is the combination of fingerprint analytical techniques with multivariate methods of one-class classifiers (OCC). In this proof of concept, we propose a novel multiblock method for OCC that emerged from the association of the common dimension (ComDim) analysis with the dual data-driven, which allows to calculate errors of misclassification based on the orthogonal and score distances with a subsequent determination of their cut-off levels. The applicability of data-driven – common dimension (DD-ComDim) analysis was verified for the authentication of diesel S10 (10 ppm of sulfur) against S500 (500 ppm of sulfur) using two low-field 1H NMR datasets: medium-resolution (MR-NMR) and time-domain NMR relaxometry (TD-NMR). The performance of the DD-ComDim (mid-level data fusion) was compared with data-driven – soft independent modeling of class analogy (DD-SIMCA) for each data set separately and concatenated (low-level data fusion). The results suggest that this novel method can improve the quality and efficiency of the model in authentication when compared to a traditional method. In addition, it demonstrated that the threshold of separation between target and non-target classes was more evident with the mid-level data fusion approach. Other applications are currently underway, intending to verify the method's applicability with other matrices and analytical techniques to confirm the reliability of the DD-ComDim.

Rapid quality assessment of Succus Bambusae oral liquid based on near infrared spectroscopy and chemometrics

In this study, a rapid quality assessment method of SBOL was established by fourier-transform near infrared (FT-NIR) spectroscopy combined with chemometric analysis. Combined with Hotelling T 2 and DModX statistics, warning limits and control limits were set for SBOL samples quality of different manufacturers. The contents of volatile components (4-hydroxybenzaldehyde, vanillin, syringaldehyde, acetylsyringone and methyl syringate) in 100 samples of SBOL were determined by gas chromatography-Mass spectrometry (GC-MS). Partial least-squares-discriminant analysis (PLS-DA), data-driven soft independent modeling of class analogy (DD-SIMCA) and random forest (RF) algorithms were used to establish the qualitative models of SBOLs based on NIR spectroscopy, which in terms of error rate and accuracy were satisfactory. Partial least squares regression (PLSR) algorithm was used to establish a quantitative calibration model. In this case, NIR showed a good performance with the effective discrimination of SBOL from different manufacturers, but the rapid quantitative model of quality markers (Q-markers) with moderate performance. • Near-infrared spectroscopy was applied for the quality control of Succus Bambusae oral liquid (SBOL). • SBOL from different manufacturers were accurately distinguished using Chemometrics. • Partial least squares models were used for the rapid determination of active compounds in SBOL with moderate performance.