A specific case in the classification of woods by FTIR and chemometric: discrimination of Fagales from Malpighiales

Fourier transform infrared (FTIR) spectroscopic data was used to classify wood samples from nine species within the Fagales and Malpighiales using a range of multivariate statistical methods. Taxonomic classification of the family Fagaceae and Betulaceae from Angiosperm Phylogenetic System Classification (APG II System) was successfully performed using supervised pattern recognition techniques. A methodology for wood sample discrimination was developed using both sapwood and heartwood samples. Ten and eight biomarkers emerged from the dataset to discriminate order and family, respectively. In the species studied FTIR in combination with multivariate analysis highlighted significant chemical differences in hemicelluloses, cellulose and guaiacyl (lignin) and shows promise as a suitable approach for wood sample classification.     

Functional nonparametric classification of wood species from thermal data

In this study, thermogravimetric (TG) and differential scanning calorimetry (DSC) curves, obtained by means of a simultaneous TG/DSC analyzer, and statistical functional nonparametric methods are used to classify different wood species. The temperature ranges, where the highest probability of correct classification is reached, are also computed. As each observation is a curve, a nonparametric functional discriminant technique based on the Bayes rule and the Nadaraya–Watson regression estimator is used. It assigns a future observation to the highest probability predefined class (supervised classification). The smoothing parameter needed in this nonparametric method is selected according to the cross-validation technique. The method proposed is applied to a sample of 49 wood items (7 per wood class) and also to classify between hardwoods and softwoods. In all the cases, the samples have been successfully classified, obtaining better results with the TG curves. The results are compared with those obtained with other nonparametric methods based on boosting algorithm. A discussion about the relation of the obtained results with the referenced wood component degradation temperature ranks is presented.     

Fruit juice authentication by <Superscript>1</Superscript>H NMR spectroscopy in combination with different chemometrics tools

To discriminate orange juice from grapefruit juice in a context of fraud prevention, ¹H NMR data were submitted to different treatments to extract informative variables which were then analysed using multivariate techniques. Averaging contiguous data points of the spectrum followed by logarithmic transformation improved the results of the data analysis. Moreover, supervised variable selection methods gave better rates of classification of the juices into the correct groups. Last, independent-component analysis gave better classification results than principal-component analysis. Hence, ICA may be an efficient chemometric tool to detect differences in the ¹H NMR spectra of similar samples, and so may be useful for authentication of foods.     

1H NMR and UV-visible data fusion for determining Sudan dyes in culinary spices

Two data fusion strategies (variable and decision level) combined with a multivariate classification approach (Partial Least Squares-Discriminant Analysis, PLS-DA) have been applied to get benefits from the synergistic effect of the information obtained from two spectroscopic techniques: UV-visible and ¹H NMR. Variable level data fusion consists of merging the spectra obtained from each spectroscopic technique in what is called “meta-spectrum” and then applying the classification technique. Decision level data fusion combines the results of individually applying the classification technique in each spectroscopic technique. Among the possible ways of combinations, we have used the fuzzy aggregation connective operators. This procedure has been applied to determine banned dyes (Sudan III and IV) in culinary spices. The results show that data fusion is an effective strategy since the classification results are better than the individual ones: between 80 and 100% for the individual techniques and between 97 and 100% with the two fusion strategies.     

Semi‐supervised linear discriminant analysis

Fisher's linear discriminant analysis is one of the most commonly used and studied classification methods in chemometrics. The method finds a projection of multivariate data into a lower dimensional space so that the groups in the data are well separated. The resulting projected values are subsequently used to classify unlabeled observations into the groups. A semi‐supervised version of Fisher's linear discriminant analysis is developed so that the unlabeled observations are also used in the model‐fitting procedure. This approach is advantageous when few labeled and many unlabeled observations are available. The semi‐supervised linear discriminant analysis method is demonstrated on a number of data sets where it is shown to yield better separation of the groups and improved classification over Fisher's linear discriminant analysis. Copyright © 2011 John Wiley & Sons, Ltd.     

Non-linear dielectric spectroscopy: antifouling and stabilisation of electrodes by a polymer coating

Non-linear dielectric spectroscopy (NLDS) has previously been shown to produce quantitative information that is indicative of the metabolic state of various organisms, by modeling the non-linear effects of their membranous enzymes on an applied oscillating electromagnetic field using supervised multivariate analysis methods. However, the instability of the characteristics of the measuring apparatus rendered the process temperamental at best in the laboratory and impractical for field use. The main practical problem, of the non-stationarity of the electrode-solution interface and the ease with which the electrode surfaces are subject to protein fouling. It is addressed by applying a thin, electrically transparent antifouling coat to the electrodes. This reduces the interminable cleaning procedures previously required to prepare the electrodes for use, increases their usable lifetime before recleaning, and also improves the precision and linearity of multivariate models on NLDS data.     

Target selection for alignment of chromatographic signals obtained using monochannel detectors

The alignment of instrumental signals, such as chromatograms, is regarded as an important step before applying multivariate chemometric techniques for data analysis. Nowadays, many alignment techniques are available and they differ in achieving their goal. They can correct peak shifts in a set of chromatograms with differing degrees of success. Almost all alignment techniques, with few exceptions [e.g., W. Yu, B. Wu, N. Lin, K. Stone, K. Williams, H. Zhao, Comput. Biol. Chem. 30 (2006) 27], require a careful choice of the target profile. The selection of a target signal is not an easy task and some difficulties related to this selection are discussed in this paper. An analysis of several simulated sets of chromatographic signals showed that the target selection can be a crucial step if the aligned signals are then used as input to unsupervised approaches, such as, e.g., principal component analysis and to supervised methods like discriminant partial least squares. Different proposals for target selection known-to-date are reviewed. As demonstrated in our study the target profile with the highest correlation coefficient among all the signals studied gave the most satisfactory results.     

Treatment of a class-in-class modelling problem in chemical pattern recognition

A supervised learning procedure is described for the detection and characterization of classes which overlap or are located within other classes. The UNEQ classifier proposed by Derde and Massart is extended by applying tests for multivariate homogeneity and homoscedasticity, principal components analysis of each class box, and Boolean-type decision rules. The extended algorithm is suitable for class-in-class modelling. The procedure is applied to chemical data for an environmental problem involving an industrial source of emission and its immission effects.     

Objective Supervised Machine Learning-Based Classification and Inference of Biological Neuronal Networks

The classification of biological neuron types and networks poses challenges to the full understanding of the human brain’s organisation and functioning. In this paper, we develop a novel objective classification model of biological neuronal morphology and electrical types and their networks, based on the attributes of neuronal communication using supervised machine learning solutions. This presents advantages compared to the existing approaches in neuroinformatics since the data related to mutual information or delay between neurons obtained from spike trains are more abundant than conventional morphological data. We constructed two open-access computational platforms of various neuronal circuits from the Blue Brain Project realistic models, named Neurpy and Neurgen. Then, we investigated how we could perform network tomography with cortical neuronal circuits for the morphological, topological and electrical classification of neurons. We extracted the simulated data of 10,000 network topology combinations with five layers, 25 morphological type (m-type) cells, and 14 electrical type (e-type) cells. We applied the data to several different classifiers (including Support Vector Machine (SVM), Decision Trees, Random Forest, and Artificial Neural Networks). We achieved accuracies of up to 70%, and the inference of biological network structures using network tomography reached up to 65% of accuracy. Objective classification of biological networks can be achieved with cascaded machine learning methods using neuron communication data. SVM methods seem to perform better amongst used techniques. Our research not only contributes to existing classification efforts but sets the road-map for future usage of brain–machine interfaces towards an in vivo objective classification of neurons as a sensing mechanism of the brain’s structure.     

Volatile compounds profiling by using proton transfer reaction‐time of flight‐mass spectrometry (PTR‐ToF‐MS). The case study of dark chocolates organoleptic differences

Direct‐injection mass spectrometry (DIMS) techniques have evolved into powerful methods to analyse volatile organic compounds (VOCs) without the need of chromatographic separation. Combined to chemometrics, they have been used in many domains to solve sample categorization issues based on volatilome determination. In this paper, different DIMS methods that have largely outperformed conventional electronic noses (e‐noses) in classification tasks are briefly reviewed, with an emphasis on food‐related applications. A particular attention is paid to proton transfer reaction mass spectrometry (PTR‐MS), and many results obtained using the powerful PTR‐time of flight‐MS (PTR‐ToF‐MS) instrument are reviewed. Data analysis and feature selection issues are also summarized and discussed. As a case study, a challenging problem of classification of dark chocolates that has been previously assessed by sensory evaluation in four distinct categories is presented. The VOC profiles of a set of 206 chocolate samples classified in the four sensory categories were analysed by PTR‐ToF‐MS. A supervised multivariate data analysis based on partial least squares regression‐discriminant analysis allowed the construction of a classification model that showed excellent prediction capability: 97% of a test set of 62 samples were correctly predicted in the sensory categories. Tentative identification of ions aided characterisation of chocolate classes. Variable selection using dedicated methods pinpointed some volatile compounds important for the discrimination of the chocolates. Among them, the CovSel method was used for the first time on PTR‐MS data resulting in a selection of 10 features that allowed a good prediction to be achieved. Finally, challenges and future needs in the field are discussed.     

Comparison of various chemometric approaches for large near infrared spectroscopic data of feed and feed products

In the present study, different multivariate regression techniques have been applied to two large near-infrared data sets of feed and feed ingredients in order to fulfil the regulations and laws that exist about the chemical composition of these products. The aim of this paper was to compare the performances of different linear and nonlinear multivariate calibration techniques: PLS, ANN and LS-SVM. The results obtained show that ANN and LS-SVM are very powerful methods for non-linearity but LS-SVM can also perform quite well in the case of linear models. Using LS-SVM an improvement of the RMS for independent test sets of 10% is obtained in average compared to ANN and of 24% compared to PLS.     

The evaluation of probabilistic classification methods : Part 2. Comparison of SIMCA,ALLOC, CLASSY and LDA

The performance of four methods for supervised probabilistic classification (LDA, SIMCA, ALLOC and CLASSY) on three types of data sets is evaluated by means of a simulation study. The methods are also applied to some practical data sets (Iris and four data sets for wines). The evaluation criterion used for discriminatory ability is the CBS (complemented Brier score) because it has some advantages over other measures. The danger of applying resubstitution evaluation for method comparison is demonstrated, but leave-one-out evaluation is shown to perform satisfactorily. Horn's method for selecting the number of principal components in SIMCA and CLASSY models is shown to be superior to the average-eigenvalue criterion. It is concluded that CLASSY is a robust method, but that in practice all the methods investigated perform about equally well on average.     

Multivariate curve resolution modeling of liquid chromatography-mass spectrometry data in a comparative study of the different endogenous metabolites behavior in two tomato cultivars treated with carbofuran pesticide

A metabonomic study based on the application of multivariate curve resolution and alternating least squares (MCR-ALS) to three-way data sets obtained by liquid chromatography coupled to mass spectrometry detection (LC-MS) was carried out for Rambo and Raf tomato cultivars treated with carbofuran pesticide. Samples were picked up during a 21 days period after treatment and analyzed by LC-MS in scan mode, along with the corresponding blank samples. Then, MCR-ALS was applied to the three-way data sets using column wise augmented matrices, and the evolutionary profiles as a function of the time after treatment were estimated for the metabolites present in both cultivars, as well as their corresponding pure spectra estimations. A comparative study using those estimations showed that some of these metabolites followed different behavior for the different cultivars after treatment. Since all treated and untreated Rambo and Raf samples were picked up according to the same sampling protocol and in a similar state of maturation, any difference in the behavior between profiles can be interpreted as an effect due to the presence of pesticide and to the kind of cultivar. Based on this hypothesis, several PLS-DA approaches were tested to check if it would be possible to classify samples by using the metabolites MCR estimations. Results showed that PLS-DA models for classification of treated or non-treated (blank) samples were the best ones obtained (98.44% of correct classifications for the validation set), which supports the stress effects related to carbofuran treatment. In addition, excellent discrimination among the four groups could be attained (89.06% of correct classifications for the validation set).     

Two-dimensional linear discriminant analysis for classification of three-way chemical data

The two-dimensional linear discriminant analysis (2D-LDA) algorithm was originally proposed in the context of face image processing for the extraction of features with maximal discriminant power. However, despite its promising performance in image processing tasks, the 2D-LDA algorithm has not yet been used in applications involving chemical data. The present paper bridges this gap by investigating the use of 2D-LDA in classification problems involving three-way spectral data. The investigation was concerned with simulated data, as well as real-life data sets involving the classification of dry-cured Parma ham according to ageing by surface autofluorescence spectrometry and the classification of edible vegetable oils according to feedstock using total synchronous fluorescence spectrometry. The results were compared with those obtained by using the spectral data with no feature extraction, U-PLS-DA (Partial Least Squares Discriminant Analysis applied to the unfolded data), and LDA employing TUCKER-3 or PARAFAC scores. In the simulated data set, all methods yielded a correct classification rate of 100%. However, in the Parma ham and vegetable oil data sets, better classification rates were obtained by using 2D-LDA (86% and 100%), compared with no feature extraction (76% and 77%), U-PLS-DA (81% and 92%), PARAFAC-LDA (76% and 86%) and TUCKER3-LDA (86% and 93%).     

Applications of maximum likelihood principal component analysis: incomplete data sets and calibration transfer

The application of a new method to the multivariate analysis of incomplete data sets is described. The new method, called maximum likelihood principal component analysis (MLPCA), is analogous to conventional principal component analysis (PCA), but incorporates measurement error variance information in the decomposition of multivariate data. Missing measurements can be handled in a reliable and simple manner by assigning large measurement uncertainties to them. The problem of missing data is pervasive in chemistry, and MLPCA is applied to three sets of experimental data to illustrate its utility. For exploratory data analysis, a data set from the analysis of archeological artifacts is used to show that the principal components extracted by MLPCA retain much of the original information even when a significant number of measurements are missing. Maximum likelihood projections of censored data can often preserve original clusters among the samples and can, through the propagation of error, indicate which samples are likely to be projected erroneously. To demonstrate its utility in modeling applications, MLPCA is also applied in the development of a model for chromatographic retention based on a data set which is only 80% complete. MLPCA can predict missing values and assign error estimates to these points. Finally, the problem of calibration transfer between instruments can be regarded as a missing data problem in which entire spectra are missing on the ‘slave’ instrument. Using NIR spectra obtained from two instruments, it is shown that spectra on the slave instrument can be predicted from a small subset of calibration transfer samples even if a different wavelength range is employed. Concentration prediction errors obtained by this approach were comparable to cross-validation errors obtained for the slave instrument when all spectra were available.     

Micro-Raman spectroscopy of natural and synthetic indigo samples

In this work indigo samples from three different sources are studied by using Raman spectroscopy: the synthetic pigment and pigments from the woad (Isatis tinctoria) and the indigo plant (Indigofera tinctoria). 21 samples were obtained from 8 suppliers; for each sample 5 Raman spectra were recorded and used for further chemometrical analysis. Principal components analysis (PCA) was performed as data reduction method before applying hierarchical cluster analysis. Linear discriminant analysis (LDA) was implemented as a non-hierarchical supervised pattern recognition method to build a classification model. In order to avoid broad-shaped interferences from the fluorescence background, the influence of 1st and 2nd derivatives on the classification was studied by using cross-validation. Although chemically identical, it is shown that Raman spectroscopy in combination with suitable chemometric methods has the potential to discriminate between synthetic and natural indigo samples.     

FTIR spectroscopy coupled with multivariate classification methods to identify different powdered infant formulas adulterated with melamine and cyanuric acid

Fourier transform infrared spectroscopy (FTIR) is a nondestructive, simple, rapid, and cheap measurement technique for analysis of many multicomponent chemical systems, e.g., detection of adulterants in food samples. In this respect, this study proposes combining FTIR spectroscopy with multivariate classification methods for classification and discrimination of different samples of infant formulas adulterated by melamine or/and cyanuric acid. Different parametric and non-parametric multivariate classification methods including the linear discriminant analysis (LDA), partial least squares-discriminant analysis (PLS-DA), soft independent modeling of class analogy (SIMCA), K-nearest neighbors (KNN), and classification and regression tree (CART) approaches were used to classify the recorded FTIR data. Assessing the performance of the multivariate methods according to their sensitivity, specificity and percent of correct prediction results demonstrated that coupling FTIR spectroscopy with multivariate classification can be applied as a rapid and powerful technique to the simultaneous detection of melamine and cyanuric acid in powdered infant formulas. This combinatorial method is efficient for adulterant concentrations as low as 0.0001 w/w%.     

Probabilistic model‐based discriminant analysis and clustering methods in chemometrics

In chemometrics, the supervised and unsupervised classification of high‐dimensional data has become a recurrent problem. Model‐based techniques for discriminant analysis and clustering are popular tools, which are renowned for their probabilistic foundations and their flexibility. However, classical model‐based techniques show a disappointing behaviour in high‐dimensional spaces, which up to now have been limited in their use within chemometrics. The recent developments in model‐based classification overcame these drawbacks and enabled the efficient classification of high‐dimensional data, even in the ‘small n / large p’ condition. This work presents a comprehensive review of these recent approaches, including regularization‐based techniques, parsimonious modelling, subspace classification methods and classification methods based on variable selection. The use of these model‐based methods is also illustrated on real‐world classification problems in chemometrics using R packages. Copyright © 2013 John Wiley & Sons, Ltd.