This Master's thesis focuses on improving the understanding and monitoring of the Hot Isostatic Pressing (HIP) sintering process used by Seco Tools AB to manufacture cemented carbides for cutting tools. While essential for producing cutting tools with superior hardness and toughness the HIP sintering process introduces a complex relationship between the selected process parameters and the achieved materials properties. With the goal of establishing batch process monitoring capabilities, this master thesis employs Multivariate Statistical Process Control (MSPC) strategies through the creation of Batch Evolution Models (BEMs) and Batch Level Models (BLMs) to monitor, predict end-product quality, and analyze the batch production HIP sintering process.  The developed models effectively account for significant variation in the HIP sintering process and demonstrate potential in identifying deviant batches. Enhancements to the models' performance are achieved through the incorporation of preprocessing, phase-specific variable selection, and specialized model training. These proposed enhancements yield discernible improvements, as evidenced by enhanced model fit and other statistical metrics.  Challenges arise when the models are tested with real-time data due to progressive changes in some tracked process variables. Block-scaling is applied to restore the real-time monitoring capabilities, but also introduces additional complexity to the models. In addition, this master thesis highlights the need for continuous and regular maintenance of these models to ensure real-time monitoring and anomaly detection capabilities. The models demonstrate varied effectiveness in predicting final product quality. For instance, they exhibit some potential in predicting Magnetic Saturation (MS), but their ability to predict Magnetic Coercivity (HC) seems nonexistent. Despite attempts to improve the predictive abilities the models are still not able to confidently predict these metrics. The master’s thesis highlights variability in powder contents and access to data of known quality nonconformities as potential areas for improving the predictive models.

No description available.

<p>The use of traditional univariate analyses for comparing groups in high-dimensional genomic studies, such as the ordinary t-test that is typically used to compare two independent groups, might be suboptimal because of methodological challenges including multiple testing problem and failure to incorporate correlation among genes. Hence, multivariate methods are preferred for the joint analysis of a group or set of variables. These methods aim to test for differences in average values of a set of variables across groups. The variables that make the set could be determined statistically (using exploratory methods such as cluster analysis) or biologically (based on membership to known pathways). In this thesis, the traditional One-Way Multivariate Analysis of Variance (MANOVA) method and a robustifed version of MANOVA (Robustifed MANOVA) are compared with respect to Type I error rates and power through a simulation study. We generated data from multivariate normal as well as multivariate gamma distributions with different parameter settings. The methods are illustrated using a real gene expression data. In addition, we investigated a popular method known as Gene Set Enrichment Analysis (GSEA), where sets of genes (variables) that belong to known biological pathways are considered jointly and assessed whether or not they are "enriched" with respect to their association with a disease or phenotype of interest. We applied this method to a real genotype data.</p> / Master of Science (MSc)

This thesis consists of three sections, the first two deal with the development and testing of an electrical impedance tomography prototype system for imaging breast cancer. The third section sues mutlivariate statistical methods to improve EIT image quality. The McMaster EIT System Mk1.0 is the resultant system of the system development. The EIT system is a 48 electrode, single current source, serial acquisition device with an operational frequency between 100Hz to 125kHz. The device is able to inject current between any two electrodes and is able to perform single or differential measurements on any two electrode pairs. The system is equipped with a virtual phase-lock loop and is capable of paramatic imaging. The system was tested using tests common to most electrical devices and specifically designed for EIT systems, to both benchmark the system and detect any errors. The testing revealed the device while able to produce viable EIT images; system suffers from a large stray capacitance. Due to stray capacitance the system injection amplitude accuracy varies with frequency and load. The system SNR is over 100dB with a 125kHz signal with a 5mA signal and compares favourably with existing EIT systems. The CMRR of the system closely tracked the published CMRR of the underlying commercial components and is comparable to existing systems. A second source of error which needs to be rectified in future deigns is the high contact impedence; which causes high direct current offset. Multivariate testing was used to detect errors which could not be easily discovered using conventional testing. The testing, performed iteratively detected several electronic errors which were fixed during development of the device. Six related models were developed for system noise, each with a different set of underlying assumptions about the source of noise. Of the models only one model proved to be a success on both qualitative and quantitative analysis of sample data sets. Finally an alternate model to the Cole-Cole parametric imaging based on PCA was proposed. The model proved to be better at modeling the underlying tissue variations in the presence of noise than Cole-Cole based models. The prototype EIT system presented in this thesis is a viable EIT system, but is in need of improvements to shielding to improve system performance. Also in need of improvement is the operational frequency and modifications toward a distributed architecture. The multivariate methods used for modelling system noise and tissue should be combined into one method for maximum benefit. / Thesis / Master of Applied Science (MASc)

The initial portion of this research focuses on appropriate parameter estimators within a general context of multivariate quality control. The goal of Phase I analysis of multivariate quality control data is to identify multivariate outliers and step changes so that the estimated control limits are sufficiently accurate for Phase II monitoring. High breakdown estimation methods based on the minimum volume ellipsoid (MVE) or the minimum covariance determinant (MCD) are well suited to detecting multivariate outliers in data. Because of the inherent difficulties in computation many algorithms have been proposed to obtain them. We consider the subsampling algorithm to obtain the MVE estimators and the FAST-MCD algorithm to obtain the MCD estimators. Previous studies have not clearly determined which of these two estimation methods is best for control chart applications. The comprehensive simulation study here gives guidance for when to use which estimator. Control limits are provided. High breakdown estimation methods such as MCD and MVE can be applied to a wide variety of multivariate quality control data. The final, lengthier portion of this research considers profile monitoring. Profile monitoring is a relatively new technique in quality control used when the product or process quality is best represented by a profile (or a curve) at each time period. The essential idea is often to model the profile via some parametric method and then monitor the estimated parameters over time to determine if there have been changes in the profiles. Because the estimated parameters may be correlated, it is convenient to monitor them using a multivariate control method such as the T-squared statistic. Previous modeling methods have not incorporated the correlation structure within the profiles. We propose the use of mixed models (both linear and nonlinear) to monitor linear and nonlinear profiles in order to account for the correlation structure within a profile. We consider various data scenarios and show using simulation when the mixed model approach is preferable to an approach that ignores the correlation structure. Our focus is on Phase I control chart applications. / Ph. D.

A rede de drenagem se constitui em importante indicador de alterações ocorridas na composição da paisagem de bacias hidrográficas, seja por mudanças na sua estruturação, forma ou então por ganho ou perda de canais. O objetivo deste trabalho é apresentar e aplicar uma proposta metodológica, com base em análise morfométrica temporal das redes de drenagem de microbacias, visando a avaliação das alterações e o zoneamento regional de uma bacia hidrográfica. A metodologia envolve o uso de imagens de sensoriamento remoto e técnicas de geoprocessamento para a caracterização das atividades antrópicas e levantamento das variáveis morfométricas em duas datas, com um intervalo de 23 anos. A avaliação das alterações e a consequente classificação das microbacias são efetuadas com base em análise estatística multivariada. A proposta foi aplicada na bacia hidrográfica do rio Capivari e os resultados comprovam que ocorreram alterações significativas na composição da rede de drenagem no período de análise. As variáveis morfométricas que mais contribuíram para a avaliação das alterações foram a Densidade de Drenagem e Densidade Hidrográfica. Dentre as microbacias que mais se alteraram estão as urbanas ou em processo de urbanização e, dentre aquelas onde as alterações foram menos pronunciadas estão as microbacias rurais, com predomínio de pastagens e/ou cana de açúcar. / The drainage system is an important indicator of the alterations occurred in the composition of the hydrographic basin environment, either due to changes in their structure and shape or due to channel loss and gain. The objective of this project is to present and apply a methodological proposal, based on time morphometric analysis of the microbasin drainage systems, aiming at the alteration evaluation and the regional zoning of a hydrographic basin. The methodology involves the use of remote images and geoprocessing techniques for the characterization of the anthropic activities and the determination of the morphometric variables on two dates, with a 23-year interval. The evaluation of the alterations and the consequent microbasin classification are carried out based on multivariate statistical analysis. The proposal was applied on Capivari river hydrographic basin, and the results prove that there were significative alterations in the drainage system composition in the analyzed period. The morphometric variables which most contributed to the alteration evaluation were the Drainage Density and the Hydrographic Density. Among the microbasins which were most altered there are the urban and under urbanization ones, and among those which had less significative alterations there are the rural microbasins, prevailing the pasture lands and sugar-cane raising.

A presente tese apresenta proposições para o uso da seleção de variáveis no aprimoramento do controle estatístico de processos multivariados (MSPC) em bateladas, a fim de contribuir com a melhoria da qualidade de processos industriais. Dessa forma, os objetivos desta tese são: (i) identificar as limitações encontradas pelos métodos MSPC no monitoramento de processos industriais; (ii) entender como métodos de seleção de variáveis são integrados para promover a melhoria do monitoramento de processos de elevada dimensionalidade; (iii) discutir sobre métodos para alinhamento e sincronização de bateladas aplicados a processos com diferentes durações; (iv) definir o método de alinhamento e sincronização mais adequado para o tratamento de dados de bateladas, visando aprimorar a construção do modelo de monitoramento na Fase I do controle estatístico de processo; (v) propor a seleção de variáveis, com propósito de classificação, prévia à construção das cartas de controle multivariadas (CCM) baseadas na análise de componentes principais (PCA) para monitorar um processo em bateladas; e (vi) validar o desempenho de detecção de falhas da carta de controle multivariada proposta em comparação às cartas tradicionais e baseadas em PCA. O desempenho do método proposto foi avaliado mediante aplicação em um estudo de caso com dados reais de um processo industrial alimentício. Os resultados obtidos demonstraram que a realização de uma seleção de variáveis prévia à construção das CCM contribuiu para reduzir eficientemente o número de variáveis a serem analisadas e superar as limitações encontradas na detecção de falhas quando bancos de elevada dimensionalidade são monitorados. Conclui-se que, ao possibilitar que CCM, amplamente utilizadas no meio industrial, sejam adequadas para banco de dados reais de elevada dimensionalidade, o método proposto agrega inovação à área de monitoramento de processos em bateladas e contribui para a geração de produtos de elevado padrão de qualidade. / This dissertation presents propositions for the use of variable selection in the improvement of multivariate statistical process control (MSPC) of batch processes, in order to contribute to the enhacement of industrial processes’ quality. There are six objectives: (i) identify MSPC limitations in industrial processes monitoring; (ii) understand how methods of variable selection are used to improve high dimensional processes monitoring; (iii) discuss about methods for alignment and synchronization of batches with different durations; (iv) define the most adequate alignment and synchronization method for batch data treatment, aiming to improve Phase I of process monitoring; (v) propose variable selection for classification prior to establishing multivariate control charts (MCC) based on principal component analysis (PCA) to monitor a batch process; and (vi) validate fault detection performance of the proposed MCC in comparison with traditional PCA-based and charts. The performance of the proposed method was evaluated in a case study using real data from an industrial food process. Results showed that performing variable selection prior to establishing MCC contributed to efficiently reduce the number of variables and overcome limitations found in fault detection when high dimensional datasets are monitored. We conclude that by improving control charts widely used in industry to accomodate high dimensional datasets the proposed method adds innovation to the area of batch process monitoring and contributes to the generation of high quality standard products.

Em um exame mamográfico, quando os fótons de raios-X incidem sobre a mama, uma parte destes fótons é transmitida sem ser desviada da trajetória inicial (radiação transmitida primária), permitindo a formação da imagem mamográfica, e outra é dispersa de sua trajetória inicial pelo tecido (radiação espalhada), atuando de forma deletéria na imagem mamográfica. Entretanto, recentes investigações têm demonstrado que a radiação espalhada pode ser útil na caracterização de tecidos. O objetivo deste trabalho é desenvolver um modelo de diagnóstico de alterações no tecido mamário utilizando as informações presentes na distribuição angular da radiação espalhada (perfil de espalhamento). Os perfis de espalhamento de 40 amostras de tecidos mamários foram obtidos utilizando um difractômetro comercial SIEMENS D5005, operando em modo reflexão na energia de 8,04 keV e variando o detector da posição angular de 5º a 150º, correspondendo a um intervalo de x de 0,03Å-1 a 0,62 Å-1. As amostras de tecido foram previamente classificadas histopatologicamente como tecidos normais, fibroadenomas (neoplasias benignas) e diferentes tipos de carcinomas (neoplasias malignas). Neste trabalho, três modelos de diagnóstico baseados na análise estatística dos perfis de espalhamento foram desenvolvidos. O primeiro, analisa seis parâmetros extraídos dos perfis de espalhamento, já o segundo e o terceiro utilizam análise multivariada (análise de componentes principais e análise de discriminante, respectivamente) para reconhecimento de padrões. Para cada modelo, valores de sensibilidade, especificidade e índice de concordância entre o diagnóstico baseado no modelo utilizado e o diagnóstico histopatológico foram obtidos. Dentre os modelos desenvolvidos, aquele que utiliza análise de discriminante proporciona o melhor diagnóstico das alterações encontradas no tecido, permitindo diferenciar tecidos normais e neoplasias benignas e malignas. Com base nos resultados obtidos conclui-se que modelos baseados na análise estatística dos perfis de espalhamento permitem classificar histologicamente tecidos mamários. / In mammography, when x-ray photons reach the breast, a fraction of these photons is transmitted without interaction with any tissues (primary transmitted radiation), allowing the formation of the mammographic image, and another fraction of them is deviated by the tissue from its initial trajectories (scattered radiation), reducing the image contrast. However, recent investigations have demonstrated that scattered radiation can be a useful diagnostic tool. The purpose of this work is to develop a diagnostic model for breast tissue characterization using the angular distribution of the scattered radiation (scattering profile). The scattering profiles of 40 breast tissue samples were obtained in a SIEMENS D5005 diffractometer, operating in reflection mode at 8,04keV, and varying the angular position of the detector from 5º to 150º, corresponding to an x interval from 0,03 Å-1 to 0,62 Å-1. All tissue samples were previously classified histopathologically as normal tissues, fibroadenomas (benign alteration) and several types of carcinomas (malignant alteration). Three models of diagnostic based on the statistical analysis of the scattering profiles were developed. The first one was constructed using six parameters extracted from the scattering profiles and the second and third models used the whole information from the scattering profiles. The latter two used multivariate analysis (principal component analysis and discriminant analysis, respectively) for pattern recognition. For each model, values of sensitivity, specificity and rate of agreement between the model diagnostic and histopathological results were obtained. Among the developed models, the discriminant analysis provides the best diagnostic of the lesions present in the tissues (normal tissues, benign and malignant alterations). From the results, it is possible to conclude that models based on the statistical analysis of the scattering profiles allow the histological classification of breast tissues.

A metabolômica vem se tornando uma abordagem eficaz para a avaliação abrangente de plantas medicinais, classificação de matérias-primas, além de estudos quimiotaxonômicos. Este trabalho demonstra a aplicabilidade da metabolômica, utilizando a subfamília Barnadesioideae (Asteraceae) como modelo de estudo, na avaliação da qualidade e classificação de espécies medicinais (espécies de Chuquiraga) e no estudo quimiotaxonômico dos principais gêneros de Barnadesioideae (Arnaldoa, Barnadesia, Chuquiraga, Dasyphyllum, Fulcaldea e Schlechtendalia). Em primeiro lugar, a análise dos perfis metabólicos por LC-MS dos membros de Barnadesioideae demonstrou que esta subfamília constitui um grupo quimicamente não explorado com uma diversidade complexa de substancias fenólicas, fenilpropanoides, alquilglicósidos e glicosídeos triterpenoides. As relações intergenéricas dentro da subfamília Barnadesioideae, baseadas na comparação dos seus perfis metabólicos por análises estatísticas multivariadas, mostraram semelhanças com as relações intergenéricas propostas pelo mais recente estudo filogenético com base em marcadores morfológicos e moleculares. Em segundo lugar, a aquisição dos perfis metabólicos de espécies de Chuquiraga (C. jussieui, C. spinosa e C. weberbaueri) por analises de LC-MS, levaram à identificação de uma variedade significativa de compostos fenólicos, fenilpropanoides, alquilglicosídeos e glicosídeos triterpenoides, assim como o estabelecimento de modelos de classificação geográfica e de espécies, além da identificação de metabólitos discriminantes por meio de análises estatísticas multivariadas exploratórias e supervisionadas. Terceiro, uma abordagem clássica foi realizada através da aquisição dos perfis cromatográficos por HPLC de espécies de Chuquiraga para o perfilhamento de compostos fenólicos e a classificação das espécies por meio de análises estatísticas multivariadas exploratórias e supervisionadas. Logo, os resultados revelam a metabolômica como uma valiosa ferramenta auxiliar no controle de qualidade e classificação de plantas medicinais, bem como em estudos de quimiotaxonômia / Metabolomics is emerging as an effective approach for the comprehensive evaluation of medicinal plants, classification of raw material, as well as chemotaxonomic studies. This work demonstrates the applicability of metabolomics, using the subfamily Barnadesioideae (Asteraceae) as a study model, for quality assessment and classification purposes of medicinal species (Chuquiraga genus) and a chemotaxonomy study of six Barnadesioideae genera (Arnaldoa, Barnadesia, Chuquiraga, Dasyphyllum, Fulcaldea and Schlechtendalia). First, the LC-MS metabolic profiles of Barnadesioideae demonstrated that this subfamily constitutes a chemically underinvestigated taxa with a complex diversity of phenolic compounds, phenylpropanoid derivatives, alkyl glycosides, and triterpenoid glycosides. The intergeneric relationships within Barnadesioideae genera, based on the comparison of their LC-MS metabolic profiles by exploratory and supervised analyses, displayed similarities to those of the intergeneric relationships obtained by the most recent phylogenetic study based on morphological and molecular markers. Second, the LC-MS metabolic profiles of three Chuquiraga species (C. jussieui, C. spinosa and C. weberbaueri) lead to the identification of a significant variety of phenolic compounds, phenylpropanoid derivatives, alkyl glycosides, and triterpenoid glycosides, as well as the establishment of prediction models for geographical origin and species classification, as well as the identification of discriminating metabolites by exploratory and supervised multivariate statistical analysis. Third, a classical approach was carried out by acquiring HPLC chromatographic profiles of three Chuquiraga species (C. jussieui, C. spinosa and C. weberbaueri) for profiling phenolic compounds and comparison by exploratory and supervised multivariate statistical analysis. Therefore, our results support metabolomics as a valuable tool in the quality control and classification of medicinal plants as well as in chemotaxonomy studies.

<p>U okviru disertacije analizirano je prisustvo različitih postojanih zagađujućih<br />materija u abiotskim i biotskim uzorcima iz različitih regiona, uključujući i uzorke<br />zemlji&scaron;ta iz Novog Sada i okolnih naselja, i to zagađujuće materije organskog<br />(policiklične aromatične ugljovodonike, polihlorovane bifenile i organohlorne<br />pesticide) i neorganskog (te&scaron;ki elementi) porekla. Dobijeni rezultati uvr&scaron;teni su u<br />baze zajedno sa relevantnim podacima iz međunarodnih radova i na taj način<br />formirane su baze koje prevazilaze lokalne interese pojedinačnih istraživanja.<br />Primenom multivarijacionih metoda analize ovakvih baza utvrđen je stepen<br />zagađenosti ispitivanih uzoraka u odnosu na rezultate iz literature, a takođe je<br />razmatrana struktura formiranih multidimenzionalnih baza sa ciljem analize<br />raspodele postojanih zagađujućih jedinjenja u posmatranim matriksima i<br />identifikacije zajedničkih izvora zagađenja. Primenom različitih (matematičkih)<br />predtretmana podataka u bazama, a zatim njihovom analizom izabranim<br />multivarijacionim metodama, izvr&scaron;ena je procena uticaja predtretmana na<br />rezultate i mogućnosti njihove interpretacije, kao i ispitivanje zavisnosti između<br />posmatranih veličina i grupisanje uzoraka. Specifični ciljevi istraživanja su<br />omogućili da se:<br />- utvrde sličnosti i razlike pri kori&scaron;ćenju različitih načina izražavanja<br />analitičkih rezultata (apsolutne vrednosti koncentracije nasuprot relativnih<br />procentualnih udela, tzv. kompozicionih podataka) u okviru baza<br />podataka i pri izdvajanju informacija iz multidimenzionalnih baza<br />primenom multivarijacionih metoda,<br />- utvrdi uticaj različitih načina pripreme (obrade) podataka pre primene<br />multivarijacionih metoda radi dobijanja potpunijih informacija u cilju bolje<br />interpretacije podataka i smanjenja dimenzija baza podataka;<br />- ispitaju regionalne i vremenske razlike i/ili sličnosti između prisustva<br />posmatranih jedinjenja u abiotskim i biotskim matriksima radi uočavanja<br />dominantnih izvora zagađenja u određenim oblastima i vremenskim<br />periodima uz istovremenu karakterizaciju eksperimentalno ispitanih<br />uzoraka u odnosu na uzorke iz drugih regiona.<br />Postignuti rezultati predstavljaju jedinstvene rezultate primene multivarijacionih<br />metoda na bazama sastavljenim od podataka dobijenim u različitim<br />istraživanjima iz sveta o prisustvu postojanih zagađujućih materija u izabranim<br />abiotskim i biotskim uzorcima, doprinoseći tako analizi njihove op&scaron;te raspodele.</p> / <p>Presence of different pollutant classes of both organic (polycyclic aromatic<br />hydrocarbons, polychlorinated biphenyls and organochlorine pesticides) and<br />inorganic origin (heavy elements) were analysed in abiotic and biotic matrices<br />from various regions, including Novi Sad and its surrounding settlements.<br />Obtained results with available data published in the international articles were<br />included in the sets, forming the input matrices to be analysed by chemometric<br />techniques. Analysis of the created sets of data by multivariate approach was<br />performed due to determining pollution level of the investigated samples, as well<br />to elucidate the persistent pollutants distribution and profiles in the selected<br />matrices and to identify the common pollution sources.<br />Using different treatments of a set of input data, influence of these procedures to<br />results was assessed.<br />Specific aims of investigation were:<br />Determination of similarity and differences by using different ways of data<br />expression (apsolute values of concentrations as apposed to relative percent<br />fraction) in interpreration of multidimension data sets on the basis of multivariate<br />statistical approach<br />Determination of different processing of data before multivariate statistical<br />methods due to obtaining adequate information for interpretation of data and<br />reducing a set of original variables<br />Examination of regional and temporal differences and/or similarity among<br />presence of observed compounds in abiotic and biotic matrices due to<br />identification of dominant pollutant sources as well as comparative<br />characterisation of experimantally obtained data in relation to samples from<br />another regions worldwide.<br />Achieved results are unique examples of multivariate methods application on<br />large data sets with results on the occurance of pesistent organic compounds in<br />abiotic and biotic matrices obtained in different studies all over the world.</p>