Unsupervised Pattern Recognition Techniques Research Articles

Discrimination/Classification of Edible Vegetable Oils from Raman Spatially Solved Fingerprints Obtained on Portable Instrumentation.

Currently, the combination of fingerprinting methodology and environmentally friendly and economical analytical instrumentation is becoming increasingly relevant in the food sector. In this study, a highly versatile portable analyser based on Spatially Offset Raman Spectroscopy (SORS) obtained fingerprints of edible vegetable oils (sunflower and olive oils), and the capability of such fingerprints (obtained quickly, reliably and without any sample treatment) to discriminate/classify the analysed samples was evaluated. After data treatment, not only unsupervised pattern recognition techniques (as HCA and PCA), but also supervised pattern recognition techniques (such as SVM, kNN and SIMCA), showed that the main effect on discrimination/classification was associated with those regions of the Raman fingerprint related to free fatty acid content, especially oleic and linoleic acid. These facts allowed the discernment of the original raw material used in the oil's production. In all the models established, reliable qualimetric parameters were obtained.

Nondestructive Metabolomic Fingerprinting: FTIR, NIR and Raman Spectroscopy in Food Screening

In recent years, there has been renewed interest in the maintenance of food quality and food safety on the basis of metabolomic fingerprinting using vibrational spectroscopy combined with multivariate chemometrics. Nontargeted spectroscopy techniques such as FTIR, NIR and Raman can provide fingerprint information for metabolomic constituents in agricultural products, natural products and foods in a high-throughput, cost-effective and rapid way. In the current review, we tried to explain the capabilities of FTIR, NIR and Raman spectroscopy techniques combined with multivariate analysis for metabolic fingerprinting and profiling. Previous contributions highlighted the considerable potential of these analytical techniques for the detection and quantification of key constituents, such as aromatic amino acids, peptides, aromatic acids, carotenoids, alcohols, terpenoids and flavonoids in the food matrices. Additionally, promising results were obtained for the identification and characterization of different microorganism species such as fungus, bacterial strains and yeasts using these techniques combined with supervised and unsupervised pattern recognition techniques. In conclusion, this review summarized the cutting-edge applications of FTIR, NIR and Raman spectroscopy techniques equipped with multivariate statistics for food analysis and foodomics in the context of metabolomic fingerprinting and profiling.

Triazine Herbicide and NPK Fertilizer Exposure: Accumulation of Heavy Metals and Rare Earth Elements, Effects on Cuticle Melanization, and Immunocompetence in the Model Species Tenebrio molitor

The increasing use of agrochemicals, including fertilizers and herbicides, has led to worrying metal contamination of soils and waters and raises serious questions about the effects of their transfer to different levels of the trophic web. Accumulation and biomagnification of essential (K, Na, Mg, Zn, Ca), nonessential (Sr, Hg, Rb, Ba, Se, Cd, Cr, Pb, As), and rare earth elements (REEs) were investigated in newly emerged adults of Tenebrio molitor exposed to field-admitted concentrations of a metribuzin-based herbicide and an NPK blend fertilizer. Chemical analyses were performed using inductively coupled plasma tandem mass spectrometry (ICP-MS/MS) supported by unsupervised pattern recognition techniques. Physiological parameters such as cuticle melanization, cellular (circulating hemocytes), and humoral (phenoloxidase enzyme activity) immune responses and mass loss were tested as exposure markers in both sexes. The results showed that NPK fertilizer application is the main cause of REE accumulation in beetles over time, besides toxic elements (Sr, Hg, Cr, Rb, Ba, Ni, Al, V, U) also present in the herbicide-treated beetles. The biomagnification of Cu and Zn suggested a high potential for food web transfer in agroecosystems. Gender differences in element concentrations suggested that males and females differ in element uptake and excretion. Differences in phenotypic traits show that exposure affects metabolic pathways involving sequestration and detoxification during the transition phase from immature-to-mature beetles, triggering a redistribution of resources between sexual maturation and immune responses. Our findings highlight the importance of setting limits for metals and REEs in herbicides and fertilizers to avoid adverse effects on species that provide ecosystem services and contribute to soil health in agroecosystems.

Fast computation of cluster validity measures for bregman divergences and benefits

Partitional clustering is one of the most relevant unsupervised learning and pattern recognition techniques. Unfortunately, one of the main drawbacks of these methodologies refer to the fact that the number of clusters is generally assumed to be known beforehand and automating its selection is not straightforward. On the same token, internal validity measures, such as the Silhouette index, Davies-Bouldin and Caliski-Harabasz measures have emerged as the standard techniques to be used when comparing the goodness of clustering results obtained via different clustering methods. These measures take into consideration both the inter and intra-cluster simmilarities and can be adapted to different metrics. Unfortunately, their used has been hindered due to their large computational complexities, which are commonly quadratic with respect to the number of instances of the data set. In this work, we show that the time complexity of computing the most popular internal validity measures can be utterly reduced by making used of the within-cluster errors and different properties of the Bregman divergences. This contribution ultimately allows us to massively speed-up the selection of an adequate number of clusters for a given data set as verified with extensive empirical comparisons.

SURFACE ROUGHNESS ASSESSMENT OF NATURAL ROCK JOINTS BASED ON AN UNSUPERVISED PATTERN RECOGNITION TECHNIQUE USING 2D PROFILES

The stability of a jointed rock mass is generally controlled by its shear strength that significantly depends on surface roughness. So far, different methods have been presented for determining surface roughness using 2D profiles. In this study, a new method based on the unsupervised pattern recognition technique using a combination of statistical, geostatistical, directional, and spectral methods for the quantification of the surface roughness will be proposed. To reach this goal, more than 10,000 profiles gathered from 92 surfaces of natural rock joints were scanned. The samples were collected from limestone cores of the Lar Dam located in the Mazandaran Province, Iran. After introducing a new spectral index, determined from the fast Fourier transform for measuring the unevenness of rough profiles, statistical, geostatistical, directional, and spectral features revealing waviness and unevenness of the 2D profiles were extracted, and a representative vector and profile for each surface were introduced through the weighted mean and median of the profile features. Principal component analysis (PCA) was utilized for finding the direction of the maximum variance of information. Then, clustering of the 92 samples was performed via K-means, and the silhouette measure was used in order to find the optimal number of clusters resulted in the creation of 13 clusters. To verify the procedure, a sample was selected in each cluster, and direct shear tests were performed on the samples. Comparing the experiments and the clustering results shows they are in good agreement. Thus, the method is an efficient tool for the quantitative recognition of surface roughness considering the waviness and unevenness of a surface.

Toward a Structural Health Monitoring Methodology for Concrete Structures under Dynamic Loads Using Embedded FBG Sensors and Strain Mapping Techniques.

A data-driven-based methodology for SHM in reinforced concrete structures using embedded fiber optic sensors and pattern recognition techniques is presented. A prototype of a reinforced concrete structure was built and instrumented in a novel fashion with FBGs bonded directly to the reinforcing steel bars, which, in turn, were embedded into the concrete structure. The structure was dynamically loaded using a shaker. Superficial positive damages were induced using bonded thin steel plates. Data for pristine and damaged states were acquired. Classifiers based on Mahalanobis’ distance of the covariance data matrix were developed for both supervised and unsupervised pattern recognition with an accuracy of up to 98%. It was demonstrated that the proposed sensing scheme in conjunction with the developed supervised and unsupervised pattern recognition techniques allows the detection of slight stiffness changes promoted by damages, even when strains are very small and the changes of these associated with the damage occurrence may seem negligible.

Using acoustic emission to assess the tensile behaviour of textile reinforced cementitious (TRC) matrix composites

The tensile behaviour of textile reinforced cementitious matrix composites (TRC) is investigated using acoustic emission monitoring. Three carbon-TRC-composites with different reinforcement ratios are tested. An unsupervised pattern recognition technique is used to classify the signals. In combination with tests on TRC-components, these clusters of signals are correlated with matrix cracking, textile–matrix interface damage, shear frictional mechanisms and textile tensile behaviour. The contribution of each mechanism during loading is quantified. Finally, the acoustic data are compared qualitatively with the evolutions of crack width (measured by Digital Image Correlation) and load-transfer length (obtained by Distributed Fibre Optic Sensing measurement). A model to predict crack width from load-transfer length is proposed and gives satisfactory results. This paper shows that the local behaviour of TRC is reflected in acoustic activity and that AE offers valuable information about damage processes in TRC-composites.

Development and evaluation of two different electronic tongues aiming to the discrimination of cutting agents found in cocaine seized samples

The screening and impurity profiling of drugs, like cocaine, is essential information that provides chemical and/or physical characterization to assist police agencies in understanding the trafficking and identifying drug origin. This work proposes to show the development and applications of two different electronic tongues (e-tongues) on the profiling study of cocaine seized samples. The developed intelligent devices' primary objective is the simple, quick, and remote cocaine classification samples based on the individual cutting agents added. The paper-based colorimetric sensor was fabricated in the lab using chromatographic paper as a substrate, wax printing to produce spot zones of reactions, a smartphone as image detection, and an editing image software to extract the chemical information through RBG values. The voltammetric e-tongue applied a boron-dopped diamond electrode to extract the cutting agents' electrochemical information through the square wave voltammetry (SWV) technique. In any case, both described sensors were coupled to chemometric tools for data analysis to construct the discrimination model. According to the objective, the unsupervised pattern recognition technique, Principal Component Analysis (PCA), was applied to test the capability of the device on individually discriminating the most common cutting agents of cocaine.

Evaluation of the effect of organic pollutants exposure on the antioxidant activity, total phenolic and total flavonoid content of lettuce (Lactuca sativa L.) using UV–Vis spectrophotometry and chemometrics

In the present contribution, the effects of different contaminants of emerging concerns (CECs), including parabens, drugs, and polycyclic aromatic hydrocarbons (PAHs) on antioxidant activity of Lactuca sativa L. in different concentration levels (10–500 µg L−1) were evaluated using ultraviolet–visible (UV–Vis) spectrophotometry combined with chemometric techniques. The extracts of lettuce samples were investigated for the antioxidant activity (AA), total phenolic content (TPC), and total flavonoid content (TFC) after 39 days of planting and 14 days of exposure. Then, the spectroscopic data were arranged in two different data matrices, including (i) the control lettuce samples and PAHs exposed samples and (ii) control samples and parabens/drugs exposed samples. The data were then analyzed using unsupervised pattern recognition techniques of principal component analysis (PCA) and hierarchical cluster analysis (HCA) and supervised pattern recognition method of partial least squares-discriminant analysis (PLS-DA). The PCA and HCA results suggested that samples belong to different groups. Then, the PLS-DA model was applied to indicate discrimination between types of organic pollutants. The PLS-DA results showed proper discrimination between control and exposed samples with an accuracy > 90%. It is concluded that the TFC, AA, and TPC values in controls were increased under the specific exposure circumstances.

One-class classification of special agroforestry Brazilian coffee using NIR spectrometry and chemometric tools

The near-infrared spectrometry combined with the one-class classification method was applied as quality control of the agroforestry-grown specialty coffee. A total of 34 samples were analyzed in this study. Spectral data were obtained using a NIR portable and different pre-treatment strategies for baseline correction were evaluated. Unsupervised pattern recognition (PCA and HCA) techniques were performed. The construction of the classification model was carried out using the dd-SIMCA algorithm with 19 samples acquired directly from producers that are recognized for the best quality control of the specialty type coffee. In order to test the model, 15 samples of non-specialty type, obtained in local markets, were evaluated. The classification model with the highest correct classification rate (CCR) scored 100% and 87% in the validation and test groups, respectively. The results demonstrated that the application of this strategy was successful in verifying the authenticity of specialty type agroforestry-grown coffee samples.

Chemometric Approaches in Questioned Documents

Questioned documents comprehend analysis of identity theft, forged signatures or texts, documents alterations and falsification of security documents or banknotes. Questions involving inks or paper require chemical analysis, and multivariate analysis or chemometrics has been an emerging tool for data evaluation and interpretation after instrumental data collection in this area. The goal of this study is to identify previous articles that applied multivariate analysis within questioned documents for forensic purposes. The search for articles was performed in four databases (Google Scholar, Science Direct, Pubmed and Scopus). Sixty studies, published in the last ten years, were selected. Thirty-four articles described pen inks analysis; fourteen articles comprehended printed documents studies; eight articles evaluated paper analysis, and four articles included banknotes analysis. Spectroscopy, mass spectrometry, chromatography, thermo gravimetric analysis and multivariate image analysis were the analytical methods applied to collect chemical data. Chemometrics methods included mainly unsupervised pattern recognition techniques, regression methods, and supervised pattern recognition techniques, amongst other methods. This review summarized and discussed multivariate analysis techniques applied in different questioned documents sub-areas, highlighting the importance of this knowledge for forensic analysts. In addition, it shows new research topics such as different printing and pen inks, papers and security documents analysis herein not included.

Development of a probabilistic early health warning system based on meteorological parameters

Among the other diseases, malaria and diarrhoea have a large disease burden in India, especially among children. Changes in rainfall and temperature patterns likely play a major role in the increased incidence of these diseases across geographical locations. This study proposes a method for probabilistic forecasting of the disease incidences in extended range time scale (2–3 weeks in advance) over India based on an unsupervised pattern recognition technique that uses meteorological parameters as inputs and which can be applied to any geographical location over India. To verify the robustness of this newly developed early warning system, detailed analysis has been made in the incidence of malaria and diarrhoea over two districts of the State of Maharashtra. It is found that the increased probabilities of high (less) rainfall, high (low) minimum temperature and low (moderate) maximum temperature are more (less) conducive for both diseases over these locations, but have different thresholds. With the categorical probabilistic forecasts of disease incidences, this early health warning system is found to be a useful tool with reasonable skill to provide the climate-health outlook about possible disease incidence at least 2 weeks in advance for any location or grid over India.

Recent Advances in Standardization of Herbal Drugs

Background:Herbal drugs play a significant role to maintain the human healthiness and to treat the ailments since the dawn of civilization. Moreover, these plants have provided many lead compounds that culminated in modern medicine. A single herb is regarded as mini-combinatorial library of phytoconstituents hence the quality control of herbal drugs in an herbal formulation is not an easy task because a number of factors impact their pharmacological efficiency and consistent therapeutic effects. Hence, to provide consistent beneficial therapeutic effects, standardized herbal products of consistent quality and purity are required.Methods:This review is based on publications obtained by a selective search in PubMed using the keywords “Standardized herbal products”, “fingerprinting”, “authentication”, “chemometric, hyphenated techniques”, “quality control of herbal drugs”, “identification”.Results:In the era of modernization, chromatographic techniques coupled with sophisticated spectroscopic analytical methods are used in estimating the authenticity, identity and characteristic of herbal products. Further, with the advancement of computer technology, chemometrics methods have become a leading tool with an unsupervised pattern recognition technique for handling multivariate data without prior knowledge about the studied samples and mines more beneficial and valuable information about the chemical entities from the raw data.Conclusion:Standardization of HDs chromatographic fingerprint is not always a perfect way to present all compounds. To assess the quality of medicinal plants, new ways are regularly being explored such as combination chemical fingerprint with biological methods, biofingerprint and metabolic fingerprint quality metrology, pharmacodynamics and export system of medicinal plants have been researched in some groups but still a significant amount of work is required to achieve a perfect system for quality evaluation of herbal drugs. Further, novel chemometric techniques have been unfolded that mines more beneficial and valuable information about the chemical entities from the raw data. So this review emphasis mainly on hyphenated techniques associated with chemometric method used in herbal drugs for identifying more valuable information and various methods for providing data, among which most commonly used techniques are chemometric resolution method and Principal Component Analysis (PCA) method.

Batch automated image processing of 2D seismic data for salt discrimination and basin-wide mapping

We have explored the technical utility of analyzing massive sets of digital 2D seismic data, collected and processed in dozens of different surveys, conducted more than 25 years ago, using batch, automated and unsupervised pattern recognition techniques to produce a basin-wide map of the top of salt. This workflow was developed for the United States portion of the Gulf of Mexico to detect top-salt boundaries on 2D poststack migrated lines. Texture-based attributes as well as novel, reflector-based attributes were used to discriminate between salt and nonsalt on each seismic line. Explicit measures of accuracy were not calculated because the data are unlabeled, but an assessment of confidence was used to score the boundaries. The depth to the top of the salt was estimated for more than 67% of the study area ([Formula: see text] or [Formula: see text]), 17% of the study area had insufficient data for processing and analysis, and 16% of the area did not meet confidence requirements for inclusion. The final results compared well with published maps of salt and the locations of salt-trapped fields. Reliable mapping of salt deeper than 6 s two-way time could not be achieved with this data set and approach because many seismic images had indistinguishable features at this depth. The computing time was greater than linear in the number of lines, but parallelization and changes in hardware configuration could reduce the run time of about three weeks to about three days.

Mechanical behaviour of jute fibre-reinforced polyester composite: Characterization of damage mechanisms using acoustic emission and microstructural observations

The present work deals with the mechanical characterization of a woven jute fabrics reinforced polyester resin composite. Two stacking sequences were considered: [0]8 and [+45/−45]2S. In-situ acoustic emission technique, X-ray computed tomography (X-ray CT) and post-mortem microscopic observations were used to characterise the damage mechanisms and to follow their evolutions during uniaxial tension. The microstructural analysis and X-ray CT revealed the following damage modes for the two lay-ups: fibre-matrix debondings which constitute the dominant mechanism, matrix crackings and fibre breakages and pull-outs. The acoustic emission data were processed using an unsupervised pattern recognition technique which combines principal components analysis and k-means optimized by a genetic algorithm. Both temporal and frequential features of acoustic emission signals were considered. The Laplacian score and dendrogram were used to determine the relevant and uncorrelated descriptors for clustering. Three clusters of events were obtained and the waveforms of each one were examined. Furthermore, the frequency contents of signals of each cluster were accurately investigated using power density spectrum and smoothed pseudo Wigner-Ville time-frequency distribution and the discrepancies between clusters are highlighted. Cluster 1 (Cl 1) characterized by signals of low frequency and intermediate amplitude, cluster 2 (Cl 2) characterized by signals of higher frequency and an amplitude similar to that of Cl 1 and cluster 3 (Cl 3) characterized by signals of higher amplitude. A correlation between the clusters and the damage mechanisms was established by means of interrupted tensile tests: Cl 1 is assigned to the matrix cracking, Cl 2 to the fibre-matrix debonding and Cl 3 to the fibre breakage and pull-out. The kinetic of evolution of each damage mode was monitored for the two stacking sequences.

Chemoprofiling and solamargine estimation from a Few Solanum species used as ‘Brihati’ and its market samples using a validated high-performance thin-layer chromatography method

An ingredient of ‘Dasamoola’ and ‘Laghupanchamoola’ group of drugs, the source of ‘Brihati’ has been controversial. Although the dried root of Solanum anguivi is considered as the source of the drug ‘Brihati’ according to the Ayurvedic Pharmacopoeia of India, closely related and morphologically similar few species like Solanum torvum, Solanum melongena, Solanum incanum, and Solanum insanum are known as its substitutes. In the present study, a high-performance thin-layer chromatography (HPTLC) method was developed and validated for the chemoprofiling and quantitative estimation of glycoalkaloid solamargine from 5 species of the genus Solanum as well as market samples. The developed method was precise, accurate, robust, specific, and linear. The results showed that S. incanum has the highest content of solamargine, followed by S. insanum. Out of the 9 market samples analyzed, solamargine was detected only in 3 samples. Unsupervised pattern recognition techniques, such as principal component analysis and hierarchical cluster analysis, were used to analyze the complex fingerprint patterns and to predict the grouping of samples. The method clearly segregated the field and market samples. Our study is the first attempt to evaluate the drug ‘Brihati’ and the market samples using HPTLC.

GC-MS and GC-FID analyses of the volatile constituents of Brachychiton rupestris and Brachychiton discolor, their biological activities and their differentiation using multivariate data analysis

The volatile constituents of the leaves and flowers of Brachychiton rupestris and Brachychiton discolor (Malvaceae) were characterized by GC-FID and GC-MS. A total of 92 compounds were identified. The major component of B. rupestris leaves was α-farnesene (34.57%) while in the flowers it was n-heptacosane (29.5%). (8Z, 11Z, 14Z)-heptadecatrienal (21.2%) in the leaves and palmitic acid (83.49%) in the flowers represent the abundant ingredients of B. discolor. Unsupervised pattern recognition techniques of multivariate data analysis exemplified by hierarchical cluster analysis (HCA) and principle component analysis (PCA) relying upon chromatographic results effectively discriminate the two related species. Also, both investigated species were assessed for their cytotoxic and antimicrobial efficacy. Results revealed that B. discolor flowers showed the highest cytotoxicity against human lung carcinoma (A-549) cell lines with IC50 equals 7.98 μg/mL, whereas B. discolor leaves exhibited the highest antimicrobial potency against all tested microorganisms namely Streptococcus pneumonia, Bacillus subtilis, Escherichia coli as well as Aspergillus fumigatus and Geotrichum candidum (MICs between 0.98 and 15.63 µg/mL) except for Salmonella typhi.

Characterization of chemical compounds susceptible to be extracted from cork by the wine using GC-MS and 1H NMR metabolomic approaches

This work presents a metabolomics study of cork by gas chromatography-mass spectrometry (GC-MS) and 1H nuclear magnetic resonance (NMR) spectroscopy to characterize compounds susceptible to be extracted from cork by the wine in an attempt to find a relationship between the content of these compounds and the geographical origin of cork. Cork from eleven geographical regions was studied, five from Portugal and six from Spain. Unsupervised pattern recognition techniques unveiled three main clusters of regions according to their chemical similarity but not related with geographical proximity. Nineteen compounds were found to be responsible for the clusters, including terpenes (trans-squalene, friedelin, camphene, trans-3-pinanone, 1-terpinen-4-ol, two sesquiterpenes), polyphenols (vescalagin, castalagin), among others (pyrogallol, glucosan, sitost-4-en-3-one, o-cymene, quinic acid, five unknowns). These preliminary results unveiled the potential for a more efficient selection of cork planks for stoppers production based on the compounds susceptible to be extracted from cork by the wine.

Delineation of Tumor Habitats based on Dynamic Contrast Enhanced MRI

Tumor heterogeneity can be elucidated by mapping subregions of the lesion with differential imaging characteristics, called habitats. Dynamic Contrast Enhanced (DCE-)MRI can depict the tumor microenvironments by identifying areas with variable perfusion and vascular permeability, since individual tumor habitats vary in the rate and magnitude of the contrast uptake and washout. Of particular interest is identifying areas of hypoxia, characterized by inadequate perfusion and hyper-permeable vasculature. An automatic procedure for delineation of tumor habitats from DCE-MRI was developed as a two-part process involving: (1) statistical testing in order to determine the number of the underlying habitats; and (2) an unsupervised pattern recognition technique to recover the temporal contrast patterns and locations of the associated habitats. The technique is examined on simulated data and DCE-MRI, obtained from prostate and brain pre-clinical cancer models, as well as clinical data from sarcoma and prostate cancer patients. The procedure successfully identified habitats previously associated with well-perfused, hypoxic and/or necrotic tumor compartments. Given the association of tumor hypoxia with more aggressive tumor phenotypes, the obtained in vivo information could impact management of cancer patients considerably.

Principal component analysis: Most favourite tool in chemometrics

Principal component analysis (PCA) is the most commonly used chemometric technique. It is an unsupervised pattern recognition technique. PCA has found applications in chemistry, biology, medicine and economics. The present work attempts to understand how PCA work and how can we interpret its results.