One-class Partial Least Squares Research Articles

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.

Multiblock data applied in organic grape juice authentication by one-class classification OC-PLS

A new strategy has been developed to enhance the assessment of the authenticity of whole grape juice within the organic class. This approach is based on the analysis of data from different analytical sources. The novel method employs a multiblock regression technique, specifically the one-class partial least squares (OC-PLS) classifier, to establish a relationship between each predictor block and the response variable. Sequential calculations are performed after orthogonalization with respect to the preceding regression scores. The proposed method has demonstrated effectiveness in detecting targeted samples. The results achieved of the best models for the test set had rates of up to 100 % sensitivity, 89 % specificity, and 83 % accuracy. To compare with the multiblock models, the DD-SIMCA method was employed, but it yielded inferior results when applied to visible data. The multiblock approach proved to be efficient in evaluating from different datasets of varied sources to classification of organic grape juice.

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.

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.

A green method for the authentication of sugarcane spirit and prediction of density and alcohol content based on near infrared spectroscopy and chemometric tools

Cachaça is a Brazilian beverage obtained from the fermentation of sugarcane juice (sugarcane spirit) and is considered one of the most consumed alcoholic beverages in the world with a strong economic impact on the northeastern Brazil, more specifically in the Brejo. This microregion produces sugarcane spirits with high quality associated to edaphoclimatic conditions. In this sense, analysis for sample authentication and quality control that uses solvent-free, environmentally friendly, rapid and non-destructive methods is advantageous for cachaça producers and production chain. Thus, in this work commercial cachaça samples using near-infrared spectroscopy (NIRS) were classified based on geographical origin using one-class classification Data-Driven in Soft Independent Modelling of Class Analogy (DD-SIMCA) and One-Class Partial Least Squares (OCPLS) and predicted quality parameters of alcohol content and density based on different chemometric algorithms. A total of 150 sugarcane spirits samples were purchased from the Brazilian retail market being 100 from Brejo and 50 from other regions of Brazil. The one-class chemometric classification model was obtained with DD-SIMCA using the Savitzky-Golay derivative with first derivative, 9-point window and 1st degree polynomial as preprocessing algorithm and sensibility was 96.70 % and specificity 100 % in the spectral range 7,290–11,726 cm−1. Satisfactory results were obtained in the model constructs for density and the chemometric model, iSPA-PLS algorithm with baseline offset as preprocessing, obtained root mean square errors of prediction (RMSEP) of 0.0011 mg/L and Relative Error of Prediction (REP) of 0.12 %. The chemometric model for alcohol content prediction used the iSPA-PLS algorithm with Savitzky-Golay derivative with first derivative, 9-point window and 1st degree polynomial as algorithm as preprocessing obtaining RMSEP and REP of 0.69 and 1.81 % (v/v), respectively. Both models used the spectral range from 7,290–11,726 cm−1. The results reflected the potential of vibrational spectroscopy coupled with chemometrics to build reliable models for identifying the geographical origin of cachaça samples for predicting quality parameters in cachaça samples.

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.

Authenticity of almond flour using handheld near infrared instruments and one class classifiers

Almonds (Prunus dulcis) are highly nutritious fruit seeds and are widely consumed, mainly in powder form (flour). With considerable economic value, this product has become a candidate for adulterations to increase profit by lowering the product cost. In the present work, methods were developed for identification of almond flour adulteration, using near infrared spectroscopy (NIR). Three different portable NIR instruments were evaluated to verify the authenticity of almond flours and the results were compared with a benchtop FT-NIR. Fifty-four almond flours of different varieties were adulterated with low-cost flours widely consumed in Brazil. Soft independent modeling of class analogies (SIMCA), Data driven SIMCA (DD-SIMCA), and One-class partial least squares (OCPLS) were used as classification methods. PLSR was employed to predict the purity of the sample. The classification results achieved 100% sensitivity and more than 95% specificity for samples with adulterant concentration of 5% (w/w) or above. The PLS models showed coefficients of determination (R²) greater than 0.90 for all models and RMSEP values between 3.2% and 4.8% for purity. The results of the multivariate models indicate that the portable NIR instruments are efficient for the identification and quantification of adulterated almond flour.

Optimization of enantiomer separation in flow-modulated comprehensive two-dimensional gas chromatography by response surface methodology coupled to artificial neural networks: Wine analysis case study

In spite of extensive applications of flow modulated comprehensive two-dimensional gas chromatography (FM-GG × GC) in different research areas, its application in the field of chiral separation is very limited. From a practical point of view, the establishment of experimental parameters for enantiomer separations is possibly more demanding in this case. Since the carrier gas flows in both dimensions, it affects not only the separation parameters, but also the fill/flush volumes of the modulator and its working efficiency. In this context, a multivariate design of experiment was applied to find the optimum experimental parameters of a reversed fill/flush (RFF) modulator for enantiomer separation of organic compounds present in botrytized wine samples. The results were described both with response surface methodology and artificial neural networks (ANN). The enantiomeric composition of chiral compounds present in the botrytized wines was used to identify their geographical origin, by principal component analysis (PCA). In addition, the developed one-class partial least squares (OC-PLS) model enabled recognition of the wine samples from the Tokaj wine region with 93% effectiveness in the presence of other samples.

Goat milk authentication by one-class classification of digital image-based fingerprint signatures: Detection of adulteration with cow milk

This work used a chemometrics-assisted color histogram-based analytical system (CACHAS) as a low-cost alternative analytical tool for the authentication of goat milk in terms of its adulteration with cow milk. For this fingerprint signatures using Grayscale RGB (red–greenblue) and HSI (hue-saturation-intensity) histograms were explored as analytical information. Then One-Class Partial Least Squares (OC-PLS) and Data-Driven Soft Independent Modeling of Class Analogy (DD-SIMCA) were employed as one-class classifiers. The best result was obtained by the RGB/DD-SIMCA model correctly classifying all pure goat milk and adulterated samples in the test set. This differentiation is favored by the presence of carotenoids in cow milk having a slightly yellowish color while goats convert β-carotene to retinol. Thus considering the complexity of the raw materials and the accessibility of the developed methodology the proposed CACHAS method stands advantageously out by finding the principles of green food analysis for screening the authenticity of goat milk.

Authentication of plant-based protein powders and classification of adulterants as whey, soy protein, and wheat using FT-NIR in tandem with OC-PLS and PLS-DA models

This study aimed at developing a non-invasive and rapid method to determine the authenticity of plant-based protein powders (free of soy, lactose, and gluten), and classify possible adulterations in the powders using near-infrared spectroscopy (NIR) and chemometric tools. Three potential powder adulterants were investigated: soy protein, whey (lactose source), and wheat (gluten source). The goal was to achieve untargeted and targeted detection to solve problems related to the authentication of the protein powders and the classification of the adulterants. For this purpose, the OC-PLS (one-class partial least squares) model was used for authentication and the PLS2-DA (partial least squares discriminant analysis) model was used to classify the adulterants. VIP (variable importance in projection) scores were used to confirm the main relevant variables and spectral ranges were responsible for each class in PLS2-DA. Laboratory samples were prepared by adding 10, 15, 20, 25, 30, 35 and 40% (w/w) of each adulterant into pure plant-based protein powder samples. In total, 47 pure plant-based protein powder samples and 144 adulterated samples were analyzed. The analysis results indicate a promising way of combining one-class (OC-PLS) with multiclass (PLS-DA) methods, in tandem with NIR to investigate plant-based protein powders. Due to the speed, high sensitivity, and specificity of the methodology, and no requirement of sample preparation, the proposed methodology could be successfully used in a range of 10–40% of adulteration, to verify the authenticity of the plant-based protein powders and to classify adulterants into soy, whey, and wheat.

In-situ authentication of goat milk in terms of its adulteration with cow milk using a low-cost portable NIR spectrophotometer

This work aimed at the use of a miniaturized near-infrared (NIR) spectrophotometer for the in-situ goat milk authentication in terms of its adulteration by adding cow milk. For this, One-Class Partial Least Squares (OC-PLS), PLS for Discriminant Analysis (PLS-DA), and the Successive Projections Algorithm for interval selection in PLS-DA (iSPA-PLS-DA) were employed. OC-PLS misclassified 1 pure goat milk sample as adulterated and 49 adulterated sample as pure in the test set, while PLS-DA misclassified 1 adulterated sample as pure goat milk and 2 pure goat milk samples as adulterated in the test set. On the other hand, the best predictive ability was achieved by iSPA-PLS-DA when using the spectra pre-processed with smoothing by 7-point window moving mean and baseline offset correction, classifying correctly 100% of the pure goat milk samples and misclassifying only one adulterated sample in the test set. This result reinforces the fact of the selection of intervals by SPA provided a more parsimonious model, with less latent variables than the other models, since it selects only the most relevant and interpretable chemical information. Therefore, the proposed methodology represents a promise, non-destructive, fast, and low-cost tool for screening the authentication of raw goat milk, also helping to support future decisions by the regulatory agencies to prevent this kind of fraud directly in the production local.

Authentication of honeybush and rooibos herbal teas based on their elemental composition

Rooibos (Aspalathus linearis) and honeybush (Cyclopia species) are popular indigenous herbal teas originating from South Africa. Both are enjoyed for their taste and aroma and more importantly, valued for their medicinal properties such as antioxidant, anti-diabetic, anti-inflammatory or immunomodulatory activity. In the European Union, rooibos and honeybush are protected as Geographical Indications. The Geographical Indication refers to products with unique characteristics that are related to their geographical origin. The authentication of products labelled as GIs is regarded as an issue of food quality and safety. Routine quality control procedures of GIs products prevent their fraud and counterfeiting on the market. However, techniques to determine adulteration or mislabelling of rooibos and honeybush do not exist yet. Therefore, in this study, the authentication of rooibos and honeybush based on their elemental composition was investigated. The methodology presented in this study combines energy-dispersive X-ray fluorescence spectrometry (EDXRF) for elemental analysis and a one-class classification approach. Elemental composition of plant material highly depends on soil on which the plant has been grown, but also on some other factors such as, e.g., atmospheric pollution, or plant metabolism. Determination of the elemental composition of the samples by EDXRF is non-destructive and does not require any complex sample preparation. One-class classification methods are well suited for authentication and origin verification problems. Based on the EDXRF data of samples from studied teas, individual class models were constructed for rooibos and honeybush. Several linear and nonlinear classification techniques were tested in order to find the model that handles the authentication task the best. For honeybush tea, the best classification results were obtained with the use of a nonlinear method based on Potential Functions. For rooibos tea, the highest classification outcomes were obtained by nonlinear One Class Partial Least Squares (OC-PLS) and the machine learning technique, Support Vector Domain Description (SVDD). The methodology implemented in the present study has the potential to be successfully applied for routine authentication of honeybush and rooibos teas.

Rapid authentication of sesame oil using ion mobility spectrometry and chemometrics

To ensure authenticity of sesame oil, an authentication technology was proposed using ion mobility spectrometry (IMS) and chemometrics. One-class classification (OCC) methods including one-class partial least squares (OCPLS) and one-class support vector machine (OCSVM) were employed to build authentication models for sesame oil. Subsequently, an independent test set was used to validate the constructed models. Validation set of 45 adulterated oils indicated that prediction correction rate of OCPLS model reached 95.6% (43 out of 45). Moreover, the complete set of sesame oils adulterated by sesame oil essence could be identified as counterfeit. Compared with previous studies, OCPLS model could work to identify untargeted adulteration. In conclusion, OCC method could effectively detect adulterated sesame oils containing as little as 10% other vegetable oils. This study provided a rapid screening method for adulterated sesame oil in market surveillance and a reference for developing authentication methods of other edible oils.

Non-Targeted Detection of Adulterants in Almond Powder Using Spectroscopic Techniques Combined with Chemometrics.

Methods that combine targeted techniques and chemometrics for analyzing food authenticity can only facilitate the detection of predefined or known adulterants, while unknown adulterants cannot be detected using such methods. Therefore, the non-targeted detection of adulterants in food products is currently in great demand. In this study, FT-IR and FT-NIR spectroscopic techniques were used in combination with non-targeted chemometric approaches, such as one-class partial least squares (OCPLS) and data-driven soft independent modeling of class analogy (DD-SIMCA), to detect adulterants in almond powder adulterated with apricot and peanut powders. The reflectance spectra of 100 pure almond powder samples from two different varieties (50 each) were collected to develop a calibration model based on each spectroscopic technique; each model was then evaluated for four independent sets of two varieties of almond powder samples adulterated with different concentrations of apricot and peanut powders. Classification using both techniques was highly sensitive, the OCPLS approach yielded 90–100% accuracy in different varieties of samples with both spectroscopic techniques, and the DD-SIMCA approach achieved the highest accuracy of 100% when used in combination with FT-IR in all validation sets. Moreover, DD-SIMCA, combined with FT-NIR, achieved a detection accuracy between 91% and 100% for the different validation sets and the misclassified samples belong to the 5% and 7% adulteration sets. These results suggest that spectroscopic techniques, combined with one-class classifiers, can be used effectively in the high-throughput screening of potential adulterants in almond powder.

Simultaneous detection of multiple frauds in kiwifruit juice by fusion of traditional and double-quantum-dots enhanced fluorescent spectroscopic techniques and chemometrics

The feasibility of simultaneous detection of multiple cheap materials in kiwifruit juice (KFJ) was studied by two fluorescent methods and chemometrics. Because the actual frauds usually involve many known and unknown adulterants, the traditional methods to detect one or more known components are insufficient to identify adulterated KFJ. Therefore, class models of pure KFJ were established using one-class partial least squares (OCPLS) to perform untargeted analysis. To enhance the detection specificity, traditional and double-quantum-dots enhanced fluorescent spectroscopic techniques were combined to characterize pure and adulterated KFJ. By data fusion and using standard normal variables (SNV) transformation, OCPLS model with sensitivity of 0.929 was obtained. The proposed method could detect adulterations of 2% (w/w) or more syrup and artificial fruit powder. The fusion of traditional and quantum dots enhanced fluorescence was demonstrated to provide a rapid and highly sensitive method for untargeted analysis of adulterated KFJ.

Detection of flaxseed oil multiple adulteration by near-infrared spectroscopy and nonlinear one class partial least squares discriminant analysis

Near-infrared (NIR) spectroscopy is widely used to detect fraudulent food products. However, NIR spectroscopy has difficulty in detecting multiple adulterations of edible oils, which is a commonly used countermeasure against adulteration detection methods. In this study, a targeted detection approach was proposed by using NIR for multiple adulteration of flaxseed oil. A variable selection method was designed to significantly reduce the number of variables and improve the accuracy of adulterant detection. After the important variables were selected by orthogonal partial least squares discriminant analysis (OPLS-DA), a one-class partial least squares (OCPLS) was used to build a detection model from a set of Gaussian radial basis functions (GRBF) that could identify flaxseed oil adulterated at a 5% level with blends of cheaper oils. This model was validated by two independent test sets. The results indicated that this model could effectively detect single, dual, or multiple adulterants with a high accuracy of 95.8% (68 out of 71). Compared with previous studies, the model built by OPLS-OCPLS provided a rapid and effective targeted detection approach for multiple adulteration of flaxseed oil.

Non-targeted Detection of Multiple Frauds in Orange Juice Using Double Water-Soluble Fluorescence Quantum Dots and Chemometrics

The feasibility of a highly sensitive “turn-off” fluorescent probe of double quantum dots (QDs) combined with chemometrics was investigated for untargeted screening of extraneous adulterants in pure orange juice (OJ), including sucrose syrup and artificial fruit powder. Pure and adulterated OJ samples were characterized by their different quenching patterns of the two separate and strong fluorescent peaks generated by the double QDs followed by chemometrics analysis. Class models of pure OJ samples (n = 117) obtained from pressing newly harvested oranges were developed using one-class partial least squares (OCPLS) based on different signal preprocessing methods, including smoothing, taking second-order derivatives (D2) and standard normal variate (SNV) transformation. As a result, D2-OCPLS model could detect at 5.0% (w/w) of sucrose syrup and 2.0% (w/w) of artificial fruit powder in pure OJ with a sensitivity (the rate of true positives) of 97.8% and specificity (rate of true negatives) of 77.0%. In conclusion, the proposed fluorescence probe with double QDs has been demonstrated to have potential for applications in rapid and sensitive screening of adulterants in OJ, which also implies promising applications to untargeted analysis of other water-soluble food samples.

Comparison of Different Multivariate Classification Methods for the Detection of Adulterations in Grape Nectars by Using Low-Field Nuclear Magnetic Resonance

Grape is the most consumed nectar in Brazil and a relatively expensive beverage. Therefore, it is susceptible to fraud by substitution with other less expensive fruit juices. Adulterations of grape nectars by the addition of apple juice, cashew juice, and mixtures of both were evaluated by using low-field nuclear magnetic resonance (LF-NMR) and supervised multivariate classification methods. Two different approaches were investigated using one-class (only unadulterated samples (UN) were modeled) and multiclass (three classes were modeled: UN, adulterated with cashew (CAS), and adulterated with apple (APP)) strategies. For the one-class approach, soft independent modeling of class analogy (SIMCA), one-class partial least squares (OCPLS), and data-driven SIMCA (DD-SIMCA) models were built. For the multiclass approach, partial least squares discriminant analysis (PLS-DA) and multiclass SIMCA models were built. The results obtained demonstrated good performances by all the one-class methods with efficiency rates higher than 93%. For the multiclass approach, the classification of samples containing only one type of adulterant presented efficiencies higher than 90% and 97% using SIMCA and PLS-DA, respectively. The classification of samples containing blends of two adulterants was satisfactory for the CAS class, but not for the APP class when applying PLS-DA. Nevertheless, multiclass SIMCA did not provide satisfactory predictions for either of these two classes.

ZnCdSe-CdTe quantum dots: A "turn-off" fluorescent probe for the detection of multiple adulterants in an herbal honey.

To enhance the power of untargeted detection, a "turn-off" fluorescent probe with double quantum dots (QDs) was developed and coupled with chemometrics for rapid detection of multiple adulterants in an herbal (Rhus chinensis Mill., RCM) honey. The double water-soluble ZnCdSe-CdTe QDs have two separate and strong fluorescent peaks, which can be quenched by honey and extraneous adulterants with varying degrees. Class models of pure RCM honey samples collected from 6 different producing areas (n = 122) were developed using one-class partial least squares (OCPLS). Four extraneous adulterants, including glucose syrup, sucrose syrup, fructose syrup, and glucose-fructose syrup were added to pure honey samples at the levels of 0.5% to 10% (w/w). As a result, the OCPLS model using the second-order derivative (D2) spectra could detect 1.0% (w/w) of different syrups in RCM honey, with a sensitivity of 0.949. The double water-soluble QDs, which can be adjusted for analysis of other water-soluble food samples, has largely extended the capability of traditional fluorescence and will provide a potentially more sensitive and specific analysis method for food frauds.

A Fast and Inexpensive Chemometric-Assisted Method to Identify Adulteration in Acai (Euterpe oleracea) Using Digital Images

This work is concerned with an analytical method for detecting acai adulteration based on digital image (DI) assisted by mean one-class classification (OCC) chemometric approaches, namely data-driven soft independent modeling of class analogy (DD-SIMCA) and one-class partial least squares (OC-PLS). In this study, two adulterants were considered, wheat flour and cassava. Digital images were acquired, in triplicate, using a webcam (WC040 MULTILASER of 5 Mp, 2G glass lenses with USB connection) in a closed wooden box with appropriate lighting and stored in JPEG format (24 bits) with a dimension of 2880 × 1620 pixels. For all the images, a central circular area was defined, used the working region to construct the frequency histograms in the color levels considering the standard RGB (red-green-blue), HSI (hue-saturation-intensity), and grayscale color models. Preliminary results obtained by principal component analysis (PCA) indicated the formation of two sample clusters (adulterated and unadulterated). On the other hand, the formation of sample clusters with respect to the type of adulterant (wheat and cassava) was not observed. OCC (DD-SIMCA and OC-PLS) models were built using eight and four factors, respectively, showing satisfactory fit. In the prediction of an external set of samples, the following results were obtained: error rate (ER) 2 and 31%, SEN 100% for both models, and specificity (SPE) 98.14 and 78.69 for DD-SIMCA and OC-PLS, respectively.