Independent Modeling Of Class Analogy Research Articles

Rapid analysis the type of customs paper using Micro-NIR spectrometers and machine learning algorithms

Quick identification of paper types for customs is extremely crucial. Although there are a variety of researches focus on the discrimination of paper, these techniques either require complex preprocessing or large-scale instruments, which are not suitable for customs environments. In this study, we predicted the type of customs paper by using a Micro-NIR spectrometer, and compared the results with Attenuated Total Reflection-Fourier Transform Infrared Spectroscopy (ATR-FTIR). Four different classification algorithms, including linear and non-linear classifiers: K-nearest neighbor (KNN), soft independent modeling of class analogy (SIMCA), partial least squares discriminant analysis (PLS-DA), and least squares-support vector machine (LS-SVM) were employed to classify the type of paper. 20 groups of datasets were selected by Monte Carlo sampling. For Micro-NIR data, the performances of KNN and LS-SVM were outstanding than SIMCA and PLS-DA, with the average accuracies 96.06% and 98.91%, respectively. The outcome of SIMCA and PLS-DA were similar, with the average accuracies 93.00% and 93.97%. Based on the standard derivation, the best stability of models was LS-SVM (1.06%), followed by PLS-DA (1.12%), KNN (1.22%) and SIMCA (3.07%). Compared with ATR-FTIR, the effects of Micro-NIR were better, which were embodies in the better KNN and SIMCA models, and the comparable LS-SVM model. The result demonstrated that the Micro-NIR combined with machine learning algorithms was an effective method to classify the type of customs paper efficiently and quickly, even better than ATR-FTIR.

E-Eye Solution for the Discrimination of Common and Niche Celery Ecotypes

Celery (Apium graveolens L.) is a well- known plant and at the basis of the culinary tradition of different populations. In Italy, several celery ecotypes, presenting unique peculiarities, are grown by small local producers, and they need to be characterized, in order to be protected and safeguarded. The present work aims at developing a fast and non-destructive method for the discrimination of a common celery (the "Elne" celery) from a typical celery of Abruzzo (Central Italy). The proposed strategy is based on the use of an e-eye tool which allows the collection of images used to infer colorgrams. Initially, a principal component analysis model was used to investigate the trends and outliers in the data. Then, the classification between the common celery (Elne class) and celery from Torricella Peligna (Torricella class) was achieved by a discriminant analysis, conducted by sequential preprocessing through orthogonalization (SPORT) and sequential and orthogonalized covariance selection (SO-CovSel) and by a class-modelling method called soft independent modelling of class analogies (SIMCAs). Among these, the highest accuracy was provided by the strategies, based on the discriminant classifiers, both of which provided a total accuracy of 82% in the external validation.

Classification of Rice Varieties Using SIMCA Applied to NIR Spectroscopic Data.

The selection of suitable rice varieties is the key to achieve high and stable yields, and the correct identification of rice varieties is the prerequisite for seed selection. In this paper, with Kenjing No.5, No.6, and No.9 as the subjects, the effectiveness of near-infrared spectroscopy (NIRS) combined with soft independent modeling of class analogy (SIMCA) in the rapid identification of rice varieties was explored. The modeling sets of Kenjing No.5, No.6, and No.9 samples were respectively used to establish a SIMCA classification model based on principal component analysis (PCA). The accuracies of the model in classifying the rice samples in the modeling set were 100, 100, and 97.5%, respectively. Then, the established SIMCA model was used to identify the rice samples in the test set. According to the experimental findings, the SIMCA analytical method achieved 100% prediction accuracy for the Kenjing No.5, Kenjing No.6, and Hongyu 001-1 samples. For the Kenjing No.9 sample, the accuracy rate was 90% with a 10% sample of Kenjing No.9 misidentified as Kenjing No.6. Therefore, the analytical method of NIRS combined with SIMCA could effectively identify the rice varieties, providing a new approach for the correct selection of planting varieties.

Detection of ternary mixtures of virgin olive oil with canola, hazelnut or safflower oils via non-targeted ATR-FTIR fingerprinting and chemometrics

Detection of extra virgin olive oil (EVOO) adulteration with ternary mixtures of undeclared foreign oils has rarely been reported in literature. The aim of the study was to develop a fit-for-purpose and highly sensitive non-targeted fingerprinting method via ATR-FTIR spectroscopy and qualitative modelling to detect ternary blends of a cheap unsaturated oil (safflower), two high in oleic acid oils (canola and hazelnut oils) at a standard adulteration level (20% v/v). An in-house spectral library for authentic EVOOs was exploited as reference data source. The 2nd derivative ATR-FTIR spectra were carefully explored for outliers throughout the multivariate calibration procedure using the Partial Least Squares Projection to Latent Structures-Discriminant Analysis (PLS-DA) and the Soft Independent Modelling of Class Analogy (SIMCA) methods. Stepwise external validation of the resulting models signified 100% sensitivity in detecting challenging cases of adulteration, i.e. with 5% canola and 15% hazelnut oils. Taken the natural variability in composition of authentic EVOOs from different producing regions and harvest year and the complexity of the studied mixtures, the overall predictive power of every resulting model was found quite high, >92%. A prospective for expanding application at lower adulteration levels i.e. 10% v/v was also evidenced.

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.

Multiple adulterants detection in turmeric powder using Vis-SWNIR hyperspectral imaging followed by multivariate curve resolution and classification techniques

In the present contribution, visible and short wavelengths of near infrared hyperspectral imaging (Vis-SWNIR-HSI) combined with different chemometric techniques is proposed as a novel technique for turmeric authentication and multiple adulterants (corn flour, rice flour, starch, wheat flour, and zedoary) detection. In this regard, twenty-three samples of turmeric were collected as whole rhizomes or powdered from seven countries and then their VIS-SWNIR hyperspectral images were obtained in 400–1000 nm using SPECIM IQ HSI device. Two multivariate resolution techniques of multivariate curve resolution-alternating least squares (MCR-ALS) and mean-field independent component analysis (MF-ICA) were used to extract pure spatial and spectral profiles of the components, and their results were compared by projecting their solutions in the area of feasible solutions (AFSs). Then, the distribution maps of turmeric component obtained using MCR-ALS and MF-ICA were used for authentication. Principal component analysis (PCA) was used to find the pattern of authentic samples using their distribution maps. Additionally, data-driven soft independent modeling of class analogy (DD-SIMCA) was employed to find boundary between authentic and adulterated turmeric samples. On this matter, good class modelling results with sensitivity of 95 % and specificity of 100 % were obtained. Finally, partial least squares-discriminant analysis (PLS-DA) was utilized for discrimination of the of adulternats and their in binary and multi-class classification modes. On this matter, PLS-DA accuracies were acceptable for binary and mixed samples which confirmed the validity of the proposed method.

Automatic hierarchical model builder

AbstractWhen building classification models of complex systems with many classes, the traditional chemometric approaches such as discriminant analysis or soft independent modeling of class analogy often fail. Some people resort to advanced deep neural network, but this is only an option if there is access to very many samples. Another alternative often used is to build hierarchical models where subclasses are sort of peeled off one or a few at a time. Such approaches often outperform classical classification as well as deep neural network on small multi‐class problems. The downside though is that it is very cumbersome to build such hierarchies of models. It requires substantial work of a skilled person. In this paper, we develop a fully automated approach for building hierarchical models and test the performance on a number of classification problems.

Authentication of fish oil (omega-3) supplements using class-oriented chemometrics and comprehensive two-dimensional gas chromatography coupled to mass spectrometry.

Food supplement authentication is an important concern worldwide due to the ascending consumption related to health benefits and its lack of effective regulation in underdeveloped countries, making it a target of fraudulent activities. In this context, this study evaluated fish oil supplements by comprehensive two-dimensional gas chromatography coupled to mass spectrometry (GC×GC-MS) to obtain fingerprints, which were used to build predictive models for automated authentication of the most popular products sold in Brazil. The authentication process relied on a one-class classifier model using data-driven soft independent modeling of class analogy (DD-SIMCA). The output of the model was a binary classifier: certified IFOS fish oils and non-certified ones - regardless of the source of adulteration. The compositional analysis showed a significant variation in the samples, which validated the need for reliable statistical models. The DD-SIMCA algorithm is still incipient in GC×GC studies, but it proved to be an excellent tool for authenticity purposes, achieving a chemometric model with a sensitivity of 100%, specificity of 98.6%, and accuracy of 99.0% for fish oil authentication. Finally, orthogonalized partial least square discriminant analysis (OPLS-DA) was used to identify the features that distinguished the groups, which ascertained the results of the DD-SIMCA model that IFOS-certified oils are positively correlated to omega-3 fatty acids, including eicosapentaenoic acid (EPA, C20:5 n-3) and docosahexaenoic acid (DHA, C22:6 n-3).

Infrared spectroscopy-based rapid determination of adulteration in commercial sheep's milk cheese via n-hexane and ethanolic extraction

This study aimed to authenticate a commercial cheese using FT-IR spectroscopy combined with chemometrics via ethanol and n-hexane extraction. Erzincan Tulum cheese (ETC) is one of the most produced and consumed cheeses in Türkiye. With an increase in production and consumption, adulteration has become a concern. Authentic ETC (n = 24) and non-authentic cheese samples (n = 22) were purchased from local markets. Ethanol and n-hexane extracts of the ETC were used for FT-IR analysis. Aliquots of the extracts were placed on a zinc selenide crystal, dried, and scanned in the mid-infrared region (4000–650 cm−1). Partial least squares-discriminant analysis (PLS-DA), soft independent modelling of class analogy (SIMCA), and conditional entropy (CE) were used as chemometric tools. PLS-DA and SIMCA provided 99% accuracy, followed by CE as low as 93.5%. Despite some disadvantages, overall, FT-IR spectroscopy with chemometrics is simple, rapid, and an inexpensive tool to detect cheese adulteration.

Confocal Raman spectroscopy assisted by chemometric tools: A green approach for classification and quantification of octyl p-methoxycinnamate in oil-in-water microemulsions

This work proposes a green analytical method based on confocal Raman spectrometry and chemometrics tools for the qualitative and quantitative analysis of oil in water microemulsions loaded with the UVB filter octyl p-methoxycinnamate (OMC). The method does not use reagents and only 10 µL of sample are needed. The analyzed microemulsion samples were synthetized in the laboratory using decaethylene glycol mono-dodecyl ether (21.9 %) as non-ionic surfactant, ethyl alcohol (7.3 %) as co-surfactant, oleic acid (1.5 %) as oil phase and water (69.3 %). A physicochemical characterization of the samples was carried out obtaining expected values for droplet size (<20 nm), polydispersity index (<0.290) and conductivity (0.04–0.07 mS cm−1), among others. Linear discriminant analysis (LDA) after selection of variables using the successive projections algorithm (SPA) and soft independent modelling of class analogy (SIMCA) were employed to classify microemulsions with different concentrations of OMC (1.0 to 10.0 %). In the case of LDA, seven Raman spectral variables were previously selected by SPA and after this SPA-LDA model resulted in one error in the prediction set achieving an accuracy of 97.8 %. The SIMCA model (α = 0.05) presented an explained variance higher 97 % using four principal components and it allowed the correct classification of 100 % of the samples (N = 15). In the quantitative analysis, partial least squares (PLS) was used to determine OMC in a range according to international legislation. The model presented optimal statistical parameters (R2 = 0.9699; RMSEP = 0.54 %) and the prediction of samples were in close agreement with HPLC method. Moreover, the greenery of the method was estimated using the AGREE metric and an optimal value of 0.85 was obtained demonstrating the proposed analytical method results environmentally friendly.

Non-targeted spatially offset Raman spectroscopy-based vanguard analytical method to authenticate spirits: White Tequilas as a case study

Adulteration and counterfeiting are ongoing problems for alcoholic drinks, including beers, wines, and spirits. To fight against them, official analytical methods need to be complemented with faster, trustworthy, non-invasive and in-situ ones, which have been named as vanguard methods, to increase the efficiency in the detection probability of truly adulterated alcoholic drinks. The analytical methodology proposed here synergistically combines a novel measurement analytical technique (spatially offset Raman spectroscopy, SORS) with chemometrics methods, i.e., principal component analysis (PCA), soft independent modeling of class analogies (SIMCA), partial least squares regression-discriminant analysis (PLS-DA), support vectors machine, (SVM) and quantitative partial least squares regression (PLSR). The applicability of the proposal is tested with Tequila to (i) differentiate among 100 % agave and mixed white packaged Tequilas, and (ii) to predict the alcoholic content. SORS spectra of 51 samples were obtained in the 300–2000 cm−1 range, from which classification and regression models were developed. The best classification performances were obtained with PLS-DA and SVM with 100 % sensitivity, specificity and overall classification rate. PLSR exposed a better trend of the samples than PCA in the exploratory analysis; and yielded predictive models capable of foreseeing alcoholic contents with average errors lower than 4 %. These results demonstrate the potential of this fast, in-situ analytical approach to be used as a vanguard analytical method to screen adulterated or counterfeited Tequilas and to assess the conformity of the alcoholic stated in the label.

Integration of NIR spectroscopy and chemometrics for authentication and quantitation of adulteration in sweet marjoram (Origanum majorana L.)

Origanum majorana L. (marjoram) has considerable promise as a folk remedy for digestive ailments in addition to its ethnomedicinal use in gynecological disorders. Due to the high demand for the botanical raw material, it is prone to intentional and inadvertent adulteration by morphologically similar herbs. In the current study, the adulteration of marjoram with its common adulterants; oregano, thyme, olive, basil and sage has been thoroughly investigated. To distinguish between marjoram and its prevalent adulterants, an integrated near infrared (NIR) spectroscopic method coupled with chemometric data analysis was developed. For exploratory pattern recognition, unsupervised multivariate models such as principal component analysis (PCA) and hierarchical cluster analysis (HCA) were implemented, followed by supervised models. Authentication of marjoram was successfully assessed using soft independent modelling of class analogy (SIMCA) with 100% sensitivity. Orthogonal projection to latent structures-discriminant analysis (OPLS-DA) modeling was employed to discriminate pure marjoram from its deliberately adulterated mixtures. Further, amounts of the adulterants were quantified in the botanical mixtures using partial least squares regression (PLS-R). The calibration and validation goodness of fit were determined to be greater than 0.9 and almost one, with a low root mean square error of prediction (RMSEP) and high ratio of performance to deviation (RPD) and range error ratio (RER) which highlight the model significant potential. Moreover, the optimal number of latent variables was found to be 5 as computed by permutation test. In the current study, the validity and reliability of the models employed to assess marjoram authenticity and purity was evidenced by both internal and external validation methods with a little to no data pre-processing.

An approach based on virtual samples for gasoline discrimination using physicochemical properties or distillation curves

Forensic cases such as fuel theft may require comparison of gasoline samples to decide whether or not gasolines seized at different locations are from the same origin. For example, a police investigation may need to identify if a gasoline sample seized from a tanker truck is the same as a sample from a filling station tank. This work proposes a novel methodology based on gasoline physicochemical properties or distillation curves enabling comparison of gasoline samples in pairs. A pair-wise comparison approach generally requires repeated measurements. Aiming to overcome the need for numerous repeated measurements, a computational approach to generate virtual samples from real gasoline measurements was developed. Measurement uncertainties of the methods were considered in this approach. The real and virtual samples were then used to compose the data set evaluated by pattern recognition techniques. Principal component analysis (PCA) was applied for exploratory studies and soft independent modeling of class analogy (SIMCA) was used for classification. Fifty-four common gasoline samples containing anhydrous ethanol fuel (27–30%v/v) collected from fuel stations located in the state of Pernambuco (Brazil) were used in this study. The 1431 possible pairs of samples were evaluated by SIMCA, with between 96% and 99% of them completely distinguishable when physicochemical assays and distillation curves were employed, respectively. These results suggest that the proposed methodology could be useful for helping in investigations requiring this type of gasoline comparison as fast as possible. Moreover, the strategy developed could also be extended to other applications that require discrimination of samples in pairs.

Use of Near-Infrared Spectroscopy to Discriminate DFD Beef and Predict Meat Quality Traits in Autochthonous Breeds.

The potential of near-infrared reflectance spectroscopy (NIRS) to discriminate Normal and DFD (dark, firm, and dry) beef and predict quality traits in 129 Longissimus thoracis (LT) samples from three Spanish purebreeds, Asturiana de los Valles (AV; n = 50), Rubia Gallega (RG; n = 37), and Retinta (RE; n = 42) was assessed. The results obtained by partial least squares-discriminant analysis (PLS-DA) indicated successful discrimination between Normal and DFD samples of meat from AV and RG (with sensitivity over 93% for both and specificity of 100 and 72%, respectively), while RE and total sample sets showed poorer results. Soft independent modelling of class analogies (SIMCA) showed 100% sensitivity for DFD meat in total, AV, RG, and RE sample sets and over 90% specificity for AV, RG, and RE, while it was very low for the total sample set (19.8%). NIRS quantitative models by partial least squares regression (PLSR) allowed reliable prediction of color parameters (CIE L*, a*, b*, hue, chroma). Results from qualitative and quantitative assays are interesting in terms of early decision making in the meat production chain to avoid economic losses and food waste.

Comparison of Portable and Benchtop Near-Infrared Spectrometers for the Detection of Citric Acid-adulterated Lime Juice: A Chemometrics Approach.

Since the incidence of food adulteration is rising, finding a rapid, accurate, precise, low-cost, user-friendly, high-throughput, ruggedized, and ideally portable method is valuable to combat food fraud. Near-infrared spectroscopy (NIRS), in combination with a chemometrics-based approach, allows potentially rapid, frequent, and in situ measurements in supply chains. This study focused on the feasibility of a benchtop Fourier-transformation-NIRS apparatus (FT-NIRS, 1000 - 2500 nm) and a portable short wave NIRS device (SW-NIRS, 740 - 1070 nm) for the discrimination of genuine and citric acid-adulterated lime juice samples in a cost-effective manner following chemometrics study. Principal component analysis (PCA) of the spectral data resulted in a noticeable distinction between genuine and adulterated samples. Wavelengths between 1100 - 1400 nm and ‎‎1550 - 1900 nm were found to be more important for the discrimination of samples for the benchtop FT-NIRS data, while variables between 950 - 1050 nm contributed significantly to the discrimination of samples based on the portable SW-NIRS data. Following partial least squares discriminant analysis (PLS-DA) as a discriminant model, standard normal variate (SNV) or multiplicative scatter correction (MSC) transformation of benchtop FT-NIRS data and SNV in combination with the second derivative transformation of portable SW-NIRS data on the training set delivered equal accuracy (94%) in the prediction of the test set. In the soft independent modeling of class analogy (SIMCA) as a class-modeling approach, the overall performances of generated models on the auto-scaled data were 98% and 94.5% for benchtop FT-NIRS and portable SW-NIRS, respectively. As a proof of concept, NIRS technology coupled with appropriate ‎multivariate classification models enables fast detection of citric acid-adulterated ‎lime juices. In addition, the promising results of portable SW-NIRS combined with SIMCA indicated its use as a screening tool for on-site analysis of lime juices at various stages of the food supply chain.

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.

Future foods protection: Supervised chemometric approaches for the determination of adulterated insects’ flours for human consumption by means of ATR-FTIR spectroscopy

In recent years, the possibility of reintroducing edible insects in the human diet in the areas where this habit had fallen in disuse has been revaluated. The reasons lie in the noticeable nutritional properties of some edible insects (generally rich in protein, vitamins, mineral elements, polyphenols and fatty acids) and their sustainable breeding which requires a moderate amount of water, and no peculiar exigencies. Consequently, insect-based snacks, insect flours, and derived products are currently spreading on the market all over the world. Although “future foods” have been a hot and debated topic in the last years, research has not kept pace with the times, and, in the literature, only a few studies on the characterization and the authentication of edible insects (and their derived products) have been published. In the light of this, the present work aims at developing a non-destructive infrared spectroscopy (IR)-based tool for detecting adulterations in insects’ flours for human consumption. Mixtures of cricket and buffalo worm flours, prepared to mimic adulterated samples, together with pure flours of the same insects, have been analysed by FT-IR spectroscopy. Eventually, classification tools for discriminating between pure and adulterated samples, i.e., Sequential Preprocessing through ORThogonalization Discriminant Analysis (SPORT-DA) and Soft Independent Modelling of Class Analogy (SIMCA), have been employed. Both discriminant and class-modelling approaches demonstrated to be suitable for the purpose. The highest correct classification rate on cricket flour was achieved by means of SPORT-DA, which correctly predicted 100% of test samples; whereas, the most accurate predictions on the buffalo worm flour data set were obtained using SIMCA, which allowed the erroneous classification of only three (over thirty) test objects.

Lab-made 3D printed electrochemical sensors coupled with chemometrics for Brazilian coffee authentication

Analytical assurance of coffees’ geographical indication (GI) authenticity is essential for producers and consumers. In this way, chemometric methods, electrochemical techniques, and 3D printed sensors become attractive to assure the coffee’s quality. These sensors are low-cost, fast, and simple, with the possibility of miniaturization and portability. Therefore, 3D printed electrodes with chemometrics were used to classify-three Brazilian coffees from regions with GI. Further, Au/Gpt-PLA electrodes with partial least squares regression were used to detect the blending of GI coffee with traditional coffee. Soft independent modelling of class analogies coupled with cyclic voltammetry had the best performance, with 91–95% accuracy, specificity of 94–100%, and 80–83% sensitivity. Furthermore, the calibration models detected and quantified traditional coffee in all three coffees from regions with GI. The detection limits ranged from 1.4 to 10% (w/w), and quantification 4.6–32%, depending on the specific coffee.

Evaluation of FT-IR spectroscopy combined with SIMCA and PLS‑DA for detection of adulterants in pistachio butter

This work scrutinized the adulteration of pistachio butter with three potential edible oils using Fourier transform infrared spectroscopy (FT-IR) and multivariate classification methods. Each of the classes, including non-adulterated samples and adulterated samples consisting of pistachio butter mixed with various concentrations of peanut oil, corn oil and sunflower oil, were classified. For this purpose, multivariate methods, including soft independent modeling of class analogy (SIMCA) and partial least squares discriminant analysis (PLS-DA), were applied to classify the FTIR data. After evaluating the model on unknown samples, the results indicated that PLS-DA was better than the SIMCA model. The classification efficiency of 98% was obtained using the PLS-DA algorithm compared to the SIMCA model with 91% efficiency for prediction. According to the results FTIR spectroscopy, accompanied by multivariate classification methods, can be used as a rapid and reliable method for classifying and predicting adulteration in pistachio butter with edible oils.

Quality classification of gasoline samples based on their aliphatic to aromatic ratio and analysis of PONA content using genetic algorithm based multivariate techniques and ATR-FTIR spectroscopy

In this study, attenuated total reflectance fourier transform mid-infrared (ATR-FTIR) spectral information was used to classify gasoline samples according to their aliphatic to aromatic ratio using data driven soft independent modeling of class analogy (dd-SIMCA). The ATR-FTIR spectral data were also used to determine the paraffin, olefin, naphthene and aromatic (PONA) contents in gasoline samples using support vector machine regression (SVMR) model. For comparison, partial least squares discriminant analysis (PLS-DA) and partial least square regression (PLSR) were also performed for classification and prediction purposes, respectively. Genetic algorithm (GA) variable selection method was applied to achieve a better separation of the classes and select characteristic variables closely related to composition differences. The results indicated that GA-dd-SIMCA and GA-SVMR can provide better classification and prediction performances by comparison with GA-PLS-DA and GA-PLSR methods, respectively. For supervised dd-SIMCA discrimination method with GA variable selection, 100% correct classification for three different classes of the gasoline was obtained, each of which has different aliphatic/aromatic ratio. In addition, the optimal GA-SVMR model showed the determination coefficient (R2) of 0.985, 0.986, 0.960 and 0.981 and root mean square error of prediction (RMSEP) were 2.050, 1.243, 2.189 and 3.324 for parrafins, olefins, naphthenes and aromatics, correspondingly. FTIR spectroscopy coupled with multivariate data analysis procedures provides rapid, efficient, and cost-effective quality assurance methods to characterize physicochemical properties of gasoline samples.