Classification of byzantine glass samples by principal component analysis

Summary Nine samples of byzantine glass classified previously by cluster analysis are classified by principal component analysis (PCA). A visual inspection of plots in coordinates of the first two principal components gives essentially the same results as cluster analysis. In addition, PCA indicates relationships among the classification variables.

---

Klassifizierung byzantinischer Glasproben durch Analyse der Hauptbestandteile

     

The morphological characterization of particles by automated scanning electron microscopy

Summary A method is described for the characterization and the classification according to shape of microscopic objects by automated SEM. The classification is performed by a hierarchical cluster analysis on a set of Fourier coefficients that are calculated from a set of radii, measured between a well defined centroid point and the contour lines of the object. This method is incorporated in existing commercial software for automated X-ray and size analysis of airborne particulate matter (PRC, Tracor Northern). Two examples demonstrate the possibilities and limitations of this method.

---

Morphologische Charakterisierung mikroskopischer Objekte mit Hilfe der Raster-Elektronenmikroskopie

Zusammenfassung Eine Methode für die Charakterisierung und Klassifizierung mikroskopischer Objekte nach ihrer Form durch automatische Raster-Elektronenmikroskopie wurde beschrieben. Die Klassifizierung wird mittels einer hierarchischen Clusteranalyse unter Verwendung eines Satzes von Fourier-Koeffizienten durchgeführt, die aus einem Satz von Radien — gemessen zwischen einem exakt definierten Mittelpunkt und den Konturlinien der Objekte — berechnet werden. Diese Methode wird in eine kommerziell erhältliche Software für automatische Röntgen- und Größenverteilungs-analyse von luftgetragenen Staubteilchen eingebaut (PRC, Tracor Northern). An Hand von Beispielen werden die Möglichkeiten und Limitierungen dieser Methode dargestellt.

     

Applications of pattern recognition

This poster illustrates the lecture on Pattern Recognition and gives recently published and unpublished examples, mainly from the laboratory from the first author. The applications concern:

- the determination of metabolic pathways of branched chain fatty acids (by clustering),

- the development of a genetic classification of meteorites (by clustering),

- the classification of cholinergic agents according to their interaction with different receptors (by clustering),

- the structure of a data set consisting of gaschromatographic profiles in samples collected in pollution monitoring stations (by factor analysis and pattern recognition),

- factors determining GLC behaviour of solutes (by factor analysis and multiple regression),

- the classification of olive oils according to geographic origin (by principal components and pattern recognition),

- the diagnosis of thyroid status (by pattern recognition).

     

基于主成分和支持向量机浓度参量同步荧光光谱油种鉴别

基于浓度参量同步荧光光谱技术,对不同溢油类型不同油源原油样品集、引入外扰相似油源样品集进行光谱数据采集,获取其浓度同步荧光光谱矩阵Concentration-Synchronous-Matrix-Fluorescence(CSMF),利用主成分分析方法对两套不同层次的原油相关样品集进行了多类分类识别。结果表明:主成分载荷图可以很好地反映各个原油相关样品在油源上的相似程度,结合支持向量机可以实现不同溢油类型及不同油源原油的准确分类,对于引入风化和海水外扰相似油源溢油样品集,两类分类区分的结果远远高于多类分类识别的结果。通过详细的主成分分析讨论,为溢油油种鉴别提供了一种利用多类分类识别,逐步缩减嫌疑样本数量,最后通过两两分类实现溢油样品准确识别的新思路。     

Comprehensive evaluation of chemical stability of Xuebijing injection based on multiwavelength chromatographic fingerprints and multivariate chemometric techniques

A strategy for monitoring and analyzing the chemical stability of Xuebijing injection (XBJ) by multiwavelength chromatographic fingerprints and multivariate classification techniques is presented in this paper. Multiwavelength chromatographic fingerprints were constructed using chromatographic data obtained at four wavelengths (260, 280, 320, and 400?nm). The raw chromatography data were preprocessed by noise reduction, baseline correction, data normalization, and interval correlation optimized shifting (icoshift). Using this method, fingerprints of 166 samples of XBJ subjected to different forced degradation conditions (irradiation, high temperature, and a range of pH values) were properly represented. Forty-one chemical components were identified using the iPeak program. In addition, the identified peak area profiling of chemical components were used for multivariate classification analysis. Principal component analysis (PCA) and Ward’s method were used to classify different XBJ degradation samples. The PCA score plot showed that XBJ degradation samples were clustered into four groups, and the results are confirmed by Ward’s method. Ten key chemical markers under different degradation conditions were found and identified by counterpropagation artificial neural networks (CP-ANN), statistical t-tests, and UPLC-Q-TOF-MS. The results suggest that the proposed strategy could be successfully applied to the comprehensive analysis of complex chemical systems.     

A small NGS–SNP panel of ancestry inference designed to distinguish African,European, East,and South Asian populations

In this study, a small set of ancestry informative SNPs was selected to differentiate African, European, East and South Asian samples, which was detected by the next-generation sequencing technology. A total of 127 Chinese Shaanxi Han individuals were collected as test samples. No statistically significant linkage disequilibrium of any pair of loci or departure from Hardy–Weinberg equilibrium of each locus was observed in the test population. To evaluate the performance of ancestry assignment using this panel, admixture analysis, principal component analysis, and likelihood ratio calculations were conducted based on the 1000 genome data and test samples. All populations were clustered into four groups, African, European, South and East Asian populations, which were consistent with their geographical origins. The pairwise fixation index (F_ST) between populations from different continental groups ranged from 0.140 to 0.621 with average 0.415, and the pairwise F_ST between populations from the same continent ranged from 0.000 to 0.056 with average 0.012. The likelihood ratio results of 125 test individuals indicated that their ancestry components were highly possible from East Asia. In conclusion, this small set of ancestry informative SNPs can be used as a reliable tool to identify and quantify ancestry components of unknown samples.     

Monitoring of water quality in South Paris district by clustering and SIMCA classification

With the aim of obtaining a monitoring tool to assess the quality of water, a multivariate statistical procedure based on cluster analysis (CA) coupled with soft independent modelling class analogy (SIMCA) algorithm, providing an effective classification method, is proposed. The experimental data set, carried out throughout the year 2004, was composed of analytical parameters from 68 water sources in a vast southwest area of Paris. Nine variables carrying the most useful information were selected and investigated (nitrate, sulphate, chloride, turbidity, conductivity, hardness, alkalinity, coliforms and Escherichia coli). Principal component analysis provided considerable data reduction, gathering in the first two principal components the majority of information representing about 92.2% of the total variance. CA grouped samples belonging to different sites, distinctly correlating them with chemical variables, and a classification model was built by SIMCA. This model was optimised and validated and then applied to a new data matrix, consisting of the parameters measured during the year 2005 from the same objects, providing a fast and accurate classification of all the samples. The most of the examined sources appeared unchanged during the 2-year period, but five sources resulted distributed in different classes, due to statistical significant changes of some characteristic analytical parameters.     

Sample classification for improved performance of PLS models applied to the quality control of deep-frying oils of different botanic origins analyzed using ATR-FTIR spectroscopy

The selection of an appropriate calibration set is a critical step in multivariate method development. In this work, the effect of using different calibration sets, based on a previous classification of unknown samples, on the partial least squares (PLS) regression model performance has been discussed. As an example, attenuated total reflection (ATR) mid-infrared spectra of deep-fried vegetable oil samples from three botanical origins (olive, sunflower, and corn oil), with increasing polymerized triacylglyceride (PTG) content induced by a deep-frying process were employed. The use of a one-class-classifier partial least squares-discriminant analysis (PLS-DA) and a rooted binary directed acyclic graph tree provided accurate oil classification. Oil samples fried without foodstuff could be classified correctly, independent of their PTG content. However, class separation of oil samples fried with foodstuff, was less evident. The combined use of double-cross model validation with permutation testing was used to validate the obtained PLS-DA classification models, confirming the results. To discuss the usefulness of the selection of an appropriate PLS calibration set, the PTG content was determined by calculating a PLS model based on the previously selected classes. In comparison to a PLS model calculated using a pooled calibration set containing samples from all classes, the root mean square error of prediction could be improved significantly using PLS models based on the selected calibration sets using PLS-DA, ranging between 1.06 and 2.91% (w/w).     

Varietal discrimination of Australian wines by means of mid-infrared spectroscopy and multivariate analysis

This study outlines the use of mid-infrared (MIR) spectroscopy combined with principal component analysis (PCA) and linear discriminant analysis (LDA) for the varietal classification of commercial red and white table wines. Three red varieties (Cabernet Sauvignon, Shiraz and Merlot) and four white varieties (Chardonnay, Riesling, Sauvignon Blanc and Viognier) were sourced from different wine regions in Australia. Wine samples were scanned in transmission on a FOSS WineScan FT 120 from wave numbers 926 to 5012 cm⁻¹. All samples were sourced from the 2006 vintage and had not been blended with any other variety or wine from other regions. Spectral data were reduced to a small number of principal components (PCs) and LDA was then performed to successfully separate the wines into the different varieties. To test the robustness of the LDA models developed for the red wines, a set of red wines scanned in 2005 were used. Correct classification of over 95% was achieved for the validation set.     

Forward selection radial basis function networks applied to bacterial classification based on MALDI-TOF-MS

Forward selection improved radial basis function (RBF) network was applied to bacterial classification based on the data obtained by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS). The classification of each bacterium cultured at different time was discussed and the effect of parameters of the RBF network was investigated. The new method involves forward selection to prevent overfitting and generalized cross-validation (GCV) was used as model selection criterion (MSC). The original data was compressed by using wavelet transformation to speed up the network training and reduce the number of variables of the original MS data. The data was normalized prior training and testing a network to define the area the neural network to be trained in, accelerate the training rate, and reduce the range the parameters to be selected in. The one-out-of-n method was used to split the data set of p samples into a training set of size p−1 and a test set of size 1. With the improved method, the classification correctness for the five bacteria discussed in the present paper are 87.5, 69.2, 80, 92.3, and 92.8%, respectively.     

Consequences of sample size, variable selection, and model validation and optimisation, for predicting classification ability from analytical data

This article discusses problems of validating classification models especially in datasets where sample sizes are small and the number of variables is large. It describes the use of percentage correctly classified (%CC) as an indicator for success of a classification model. For small datasets, %CC should not be used uncritically and its interpretation depends on sample size. It illustrates the use of a common classification method, discriminant partial least squares (D-PLS) on a randomly generated dataset of 200 samples and 200 variables.

An aim of the classifier is to determine whether the null hypothesis (there is no distinction between two classes) can be rejected. Autoprediction gives an 84.5% CC. It is shown that, if there is variable selection, it must be performed independently on the training set to obtain a CC close to 50% on the test set; otherwise, over-optimistic and false conclusions can be reached about the ability to classify samples into groups.

Finally, two aims of determining the quality of a model are frequently confused, namely optimisation (often used to determine the most appropriate number of components in a model) and independent validation; to overcome this, the data should be split into three groups.

There are often difficulties with model building if validation and optimisation have been done on different groups of samples, especially using iterative methods, each group being modelled using properties, such as a different number of components or different variables.     

基于分步相关成分分析的中药材质量鉴别神经元分类器

提出并构建了一种基于分步相关成分分析的神经元分类器(SCCA-HBP),并将其用于中药材质量模式分类.通过从色谱分析所得到的高维数据集中分步提取分类相关成分,获取化学模式特征向量,使神经元分类器输入模式向量的维数降低.此外,提出用带输出误差死区的混合BP算法训练神经元分类器,提高了网络学习训练速度和分类准确性.以32个当归样品质量等级分类鉴别为例考察本方法,分类正确率为100%,优于PCA-BP(84.4%)和SCCA-BP(90.6%)方法;且训练时间仅为BP算法的54.2%.     

Chemometric analysis of diesel fuel for forensic and environmental applications

Diesel fuel samples were analyzed using gas chromatography-mass spectrometry (GC-MS) and chemometric procedures to associate and discriminate samples for potential use in forensic and environmental applications. Twenty-five diesel samples, representing 13 different brands, were collected from service stations in the Lansing, Michigan area. From the GC-MS data, mass-to-charge ratios were identified to represent aliphatic (m/z 57) and aromatic (m/z 91 and 141) compounds. The total ion chromatogram (TIC) and extracted ion chromatograms (EICs) of the chosen ions were evaluated using Pearson product moment correlation (PPMC) and principal component analysis (PCA). Diesel samples from the same brand showed higher PPMC coefficients, while those from different brands showed lower values. EICs generally provided a wider range of correlation coefficients than the TIC, with correspondingly increased discrimination among samples for EIC m/z 91. PCA grouped the diesel samples into four distinct clusters for the TIC. The first cluster consisted of four samples from the same brand, two clusters contained one diesel sample each of different brands, and the fourth cluster contained the remaining diesel samples. The same trend was observed using each EIC, with an increase in the number of clusters formed for EIC m/z 57 and 91. Both statistical procedures suggest aromatic components (specifically, those with m/z 91) provide the greatest discrimination among diesel samples. This conclusion was supported by identifying the chemical components that contribute the most to the variance. The relative amount of aliphatic versus aromatic components was found to cause the greatest discrimination among samples in the data set.     

Determination of cadmium, lead and copper in wine by differential pulse anodic stripping voltammetry

Summary The simultaneous determination of cadmium, lead and copper in wine by differential pulse anodic stripping voltammetry at the hanging mercury drop electrode is described. The wine samples are decomposed in a mixture of sulphuric acid and hydrogen peroxide at 180° C. The procedure is controlled by recovery tests and compared with other wet digestion methods. The results for ten red and white wine samples of different origin are given. The lead values (65–230 ppb) were below the accepted maximum level for this metal, but some of the wines contained relatively large amounts of copper (0.08–1.04 ppm). Very low values (1.4–6.6 ppb) were found for cadmium.

---

Bestimmung von Cadmium, Blei und Kupfer in Wein durch Differentialpulse-anodic-stripping Voltammetrie

Zusammenfassung Bei dem beschriebenen Verfahren werden die Proben mit Schwefelsäure/Wasserstoffperoxid bei 180° C aufgeschlossen. Recovery Tests und Vergleiche mit anderen Naßaufschluß-methoden wurden durchgeführt. Ergebnisse für 10 Proben von Rot- und Weißweinen verschiedenen Ursprungs werden angegeben. Die Bleigehalte (65–230 ppb) lagen unter den zugelassenen Maximalwerten. Einige Weine wiesen jedoch relativ hohe Kupfergehalte auf (0,08–1,04 ppm). Cadmium wurde nur in sehr geringen Mengen gefunden (1,4–6,6 ppb).

We would like to thank A/S Vinmonopolet for supplying the wine samples, and the Royal Norwegian Council for Scientific and Industrial Research for a postdoctoral fellowship (M. Oehme).     

The use of a combined portable X ray fluorescence and multivariate statistical methods to assess a validated macroscopic rock samples classification in an ore exploration survey

The combination of “ex situ” portable X ray fluorescence with unsupervised and supervised pattern recognition techniques such as hierarchical cluster analysis, principal components analysis, factor analysis and linear discriminant analysis have been applied to rock samples, in order to validate a “in situ” macroscopic rock samples classification of samples collected in the Boris Angelo mining area (Central Chile), during a drill-hole survey carried out to evaluate the economic potential of this Cu deposit. The analysed elements were Ca, Cu, Fe, K, Mn, Pb, Rb, Sr, Ti and Zn. The statistical treatment of the geological data has been arisen from the application of the Box-Cox transformation used to transform the data set in normal form to minimize the non-normal distribution of the data. From the statistical results obtained it can be concluded that the macroscopic classification applied to the transformed data permits at least, to distinguish quite well in relation to two of the rock classes defined (70.5% correctly classified (p < 0.05)) as well as for four of the five alteration types defined “in situ” (75% of the total samples).     

NIR analysis of pharmaceutical samples without reference data: improving the calibration

Using an appropriate set of samples to construct the calibration set is crucial with a view to ensuring accurate multivariate calibration of NIR spectroscopic data. In this work, we developed and optimized a new methodology for incorporating physical variability in pharmaceutical production based on the NIR spectrum for the process. Such a spectrum contains the spectral changes caused by each treatment applied to the component mixture during the production process. The proposed methodology involves adding a set of process spectra (viz. difference spectra between those for production tablets and a laboratory mixture of identical nominal composition) to the set of laboratory samples, which span the wanted concentration range, in order to construct a calibration set incorporating all physical changes undergone by the samples in each step of the production process. The best calibration model among those tested was selected by establishing the influence of spectral pretreatments used to obtain the process spectrum and construct the calibration models, and also by determining the multiplying factor m to be applied to the process spectra in order to ensure incorporation of all variability sources into the calibration model. The specific samples to be included in the calibration set were selected by principal component analysis (PCA). To this end, the new methodology for constructing calibration sets for determining the Active Principle Ingredients (API) and excipients was applied to Irbesartan tablets and validated by application to the API and excipients of paracetamol tablets. The proposed methodology provides simple, robust calibration models for determining the different components of a pharmaceutical formulation.     

Gas-chromatographische Profilanalyse der polycyclischen aromatischen Kohlenwasserstoffe in Kl?rschlammproben

Zusammenfassung Es wird ein einfaches und selektives Anreicherungsverfahren für polycyclische aromatische Verbindungen angegeben (flüssig-flüssig-Verteilung: Dimethylformamid + Wasser + Cyclohexan; Chromatographie an Sephadex LH 20/Isopropanol; Filtration an Aluminiumoxid/Cyclohexan). Das Gemisch der polycyclischen aromatischen Kohlenwasserstoffe (= PAH) wird gas-chromatographisch an gepackten Hochleistungssäulen getrennt und die FID-Signale mit einem anfangs zugegebenen inneren Standard verglichen. Es wurden 18 Hauptkomponenten quantitativ ausgewertet. Die Variationskoeffizienten einer fünffachen Bestimmung derselben Probe lagen zwischen 1,6 und 11,3%. Die Nachweisgrenze des Verfahrens beträgt bei einer mittleren elektronischen Signalverstärkung in Abhängigkeit von der Retentionszeit 0,5–5 ng (für Benzo(a)pyren 1 ng). Die Charakterisierung der PAH mit 4–7 Ringen erfolgte mit einer GC-MS-Kombination, was für den Routinebetrieb aufgrund der großen Ähnlichkeit der PAH-Profile verschiedener Klärschlammproben nicht erforderlich ist.

---

Profile-analysis of polycyclic aromatic hydrocarbons in sewage sludge by gas chromatography

Summary A simple and selective method of enrichment of polycylic aromatic compounds is described (liquid-liquid-distribution with DMF + water + Cyclohexane; chromatography on Sephadex LH 20/isopropanol; filtration on aluminium oxide/cyclohexane). The fraction of polycyclic aromatic hydrocarbons (= PAH), separated gas-chromatographically using high performance columns, is evaluated by comparising the FID-signals with those of the internal standard added to the sample. Eightteen main components are evaluated. The variation coefficient of five analyses of the sample is in the range of 1.6–11.3%. The detection limit by moderate amplification depends on retention time of the PAH (0.5–5 ng; for benzo(a)pyrene 1 ng). As the PAH profiles of different samples of sludge are very similar, it is normally not necessary to characterize the PAH by mass spectrometry.

     

Experimental examination of the matrix effect in the flame emission spectrometry of K, Na, and Sr

Summary The interferences in the determination of K, Na and Sr by flame emission spectrometry due to different sample components have been studied according to a new methodology of investigation of the matrix effect. This methodology comprises the formulation of mathematical (polynomial) models approximating the relationship between a measured signal (e.g. emission) and the concentration of the sample components. The regression coefficients in these models are estimated on the basis of the results of measurements carried out on appropriate standard samples. Their composition results from such experimental plans as a 2ⁿ factorial, a 3ⁿ factorial and a rotatable composite design. The 3ⁿ factorial one was the basis for the formulation of high-degree polynomial models which appeared adequate in all the cases considered. Some general problems connected with the use of this methodology in flame emission spectrometry (FES) are also discussed.

---

Experimentelle Untersuchung des Matrixeffektes bei der Flammenemissions-Spektrometrie von K, Na und Sr

Zusammenfassung Mit Hilfe einer neuen Untersuchungsmethodik wurden die störenden Einflüsse von Begleitsubstanzen auf die flammenphotometrische Bestimmung von K, Na und Sr untersucht. Das Verfahren besteht in der Aufstellung mathematischer (Polynom-) Modelle, die der Beziehung zwischen Meßsignal (z. B. Emission) und Konzentration des betreffenden Bestandteils angenähert sind. Die Regressionskoeffizienten in diesen Modellen wurden aufgrund von Messungen an geeigneten Standardproben berechnet. Deren Zusammensetzung ergab sich aus entsprechenden Versuchsplänen (2ⁿ- und 3ⁿ-faktoriell sowie ein Rotationsplan). Der 3ⁿ-faktorielle Plan diente als Grundlage zur Aufstellung von Polynom-Modellen höheren Grades, die allen untersuchten Fällen genügten. Einige allgemeinere Probleme in Verbindung mit der Anwendung dieser Methodik auf die Flammenphotometrie werden ebenfalls diskutiert.

     

Direct coupling of a gas-liquid separator to an ion mobility spectrometer for the classification of different white wines using chemometrics tools

The potential of a vanguard technique as is the ion mobility spectrometry with ultraviolet ionization (UV-IMS) coupled to a continuous flow system (CFS) have been demonstrated in this work using a gas phase separator (GPS). This vanguard system (CFS-GPS-UV-IMS) has been used for the analysis of different types of white wines to obtain a characteristic profile for each type of wine and their posterior classification using different chemometric tools. Precision of the method was 3.1% expressed as relative standard deviation. A deep chemometric study was carried out for the classification of the four types of wines selected. The best classification performance was obtained by first reducing the data dimensionality by principal component analysis (PCA) followed by linear discriminant analysis (LDA) and finally using a k-nearest neighbour (kNN) classifier. The classification rate in an independent validation set was 92.0% classification rate value with confidence interval [89.0%, 95.0%] at 95% confidence level.The same white wines analyzed using CFS-GPS-UV-IMS were analyzed using gas chromatography with a flame detector (GC-FID) as conventional technique. The chromatographic method used for the determination of superior alcohols in wine samples shown in the Regulation CEE 1238/1992 was selected to carry out the analysis of the same samples set and later the classification using appropriate chemometrics tools. In this case, strategies PCA-LDA and kNN classifier were also used for the correct classification of the wine samples. This combination showed similar results to the ones obtained with the proposed method.     

Entwicklung eines Testverfahrens auf Mangandioxid-Basis zur Ermittlung des Remobilisierungsverm?gens von W?ssern

Zusammenfassung Es wird über die Entwicklung eines Verfahrens zur Bestimmung des Remobilisierungsvermögens von Wässern auf MnO₂/Cu-Basis berichtet. Ferner werden Erfahrungen mit diesem Verfahren an ausgewählten bekannten Komplexbildnern unter Berücksichtigung des Einflusses von verschiedenen Wasserinhaltsstoffen mitgeteilt. Darüber hinaus werden erste Ergebnisse über das Remobilisierungsvermögen von Umweltproben vorgestellt.

---

Determination of the metal-mobilization capacity of waters employing a solid phase of MnO₂

Summary The method is based on a solid phase of MnO₂, charged with Cu, which is formed under defined conditions. The mobilization capacities of several complexing agents were tested, considering also the influence of different water components (buffer capacities, salt effects etc.). Results from real water samples are presented.

Die Untersuchungen wurden aus Mitteln des Kuratoriums für Wasserwirtschaft (KfW) und des Bundesministeriums des Innern gefördert