Nanoparticles, i.e. particles with at least one dimension smaller than 100 nm, are present in large quantities in ambient air and can also be found in an increasing amount of consumer products. It is known that many nanomaterials have physicochemical properties that differ from physicochemical properties of the same material in bulk size. It is therefore important to characterize nanoparticles and to evaluate their toxicity. To understand mechanisms behind nanotoxicity, it is important to study the uptake of nanoparticles, and how they are accumulated. For these purposes model studies of cellular uptake are useful. In this thesis metal oxide and carbon-based nanoparticles have been studied in living cells using Raman spectroscopy. Raman spectroscopy is a method that facilitates a non-destructive analysis without using any fluorescent labels, or any other specific sample preparation. It is possible to collect Raman images, i.e. images where each pixel corresponds to a Raman spectrum, and to use the spectral information to detect nanoparticles, and to identify organelles in cells. In this thesis the question whether or not nanoparticles can enter the cell nucleus of lung epithelial cells has been addressed using hyperspectral analysis. It is shown that titanium dioxide nanoparticles and iron oxide nanoparticles are taken up by cells, and also in the cell nucleus. In contrast, graphene oxide nanoparticles are mainly found attached on the outside of the cell membrane and very few nanoparticles are found in the cell, and none have been detected in the nucleus. It is concluded that graphene oxide nanoparticles are not cytotoxic. However, a comparison of Raman spectra of biomolecules in cells exposed to graphene oxide, unexposed cells and apoptotic cells, shows that the graphene oxide nanoparticles do affect lipid and protein structures. In this thesis, several multivariate data analysis methods have been used to analyze Raman spectra and Raman images. In addition, super-resolution algorithms, which originally have been developed to improve the resolution in photographic images, were optimized and applied to Raman images of cells exposed to submicron polystyrene particles in living cells.

Thesis (MSc)--Stellenbosch University, 2015. / ENGLISH ABSTRACT: The alcoholic beverage industry needs objective analysis of grape juice and liqueur quality. Fourier transform infrared (FT-IR) spectroscopy with multivariate data analysis techniques is widely used in wine laboratories across South Africa for accurate, fast and high sample throughput analyses. In this study the potential of FT-IR spectroscopy is evaluated for the quantification of ammonia in freshly pressed grape juice. FT-IR spectroscopy is evaluated, using two different spectrometers, in respectively attenuated total reflection (ATR) - and transmission scanning modes for the quantification of alcohol, pH and invert sugar in spirit-based liqueurs. The ultimate aim was to implement the PLS regression algorithms developed at an industrial cellar and replace the complex and lengthy reference methods used at the time of this study. Principle component analysis (PCA) was performed prior to the calibration step to identify groupings and patterns within the spectra. The PLS calibration models were developed from samples collected at the cellar and using partial least square (PLS) regression. The models were evaluated using the performance criteria coefficient of determination (R2) and root mean squared error of cross validation (RMSECV) at calibration stage, and root mean square error of prediction (RMSEP) and residual predictive deviation ratio (RPD) at validation stage. The average RMSEP (1.88 mg/L) of the ammonia PLS calibration model was in agreement with the standard error of laboratory (SEL = 1.54 mg/L). The R2 (92.05) and average RPD (3.3) proposed a model with excellent precision for screening purposes that was ready to be transferred for use by the laboratory. The r2 values for the alcohol, pH and invert sugar PLS calibration models obtained in ATR and transmission, indicated good to excellent precision (80<r2<100). The alcohol PLS calibration model obtained in transmission was suitable for quality- and process control purposes (RPD = 21.2), while the invert sugar PLS calibration model for quality control purposes (5<RPD<6.4). The pH and invert sugar calibration models obtained in ATR were suitable for screening purposes with RPD = 3.6 and RPD = 4.8, respectively. These PLS regression algorithms were implemented at the cellar. The pH PLS calibration model obtained in transmission was suitable for rough screening of samples (RPD = 2.7) and future development was neccesary to increase the predictability of the model. The results obtained in this study made a significant contribution towards validation of FT-MIR as a powerful tool for rapid quantification of quality indicating parameters in wine and spirit-based liqueurs. The contribution is particularly valuable in the context of ongoing research to improve the quality of products at the cellar to meet consumer demands. The knowledge gained on quantification of quality indication parameters of spirit-based liqueurs is novel and this is one of the first reports on implementation of mid-infrared (MIR) spectroscopy for the quality control of South African spirit-based liqueurs. / AFRIKAANSE OPSOMMING: Die wynindustrie benodig objektiewe analises van druiwesap- en likeurgehalte. Fourier-transformasie- infrarooi (FT-IR) spektroskopie met multiveranderlike statistiese metodes word gebruik in wynlaboratoriums regoor Suid-Afrika vir akkurate, vinnige en hoë monsterdeurset ontledings. In hierdie studie is die potensiaal van FT-IR spektroskopie geëvalueer vir die kwantifisering van ammoniak in die sap van vars geparste wyndruiwe. Twee verskillende FT-IR spektroskopie instrumente, in onderskeidelik (verswakte totale refleksie, ATR) - en transmissie skandering is gebruik vir die kwantifisering van alkohol, pH en invertsuiker in spiritus-gebaseerde likeurs. Die uiteindelike doel was om die parsiële kleinste kwadraat (PKK)- regressie algoritmes wat ontwikkel is, by 'n industriële kelder te implementeer en die komplekse en tydrowende verwysingmetodes wat tydens die studie in die kelder gebruik is te vervang. Verskeie multiveranderlike hoofkomponentanalise (MVK) is uitgevoer voor die kalibrasie stap, met die doel om groeperings en patrone in die spektra te identifiseer. Die PKK kalibrasiemodelle is ontwikkel van monsters wat by die kelder versamel is en die spektra is gebruik in die PKK regressies. Tydens die kalibrasiefase is die modelle geëvalueer met behulp van die bepalingskoëffisiënt (R2) en gemiddelde kalibrasieprediksiefout en tydens die validasiefase, met behulp van die standaardvoorspellingsfout (SVF) en relatiewe voorspellingsafwyking (RVA). Die gemiddelde SVF (1.88 mg/L) van die ammoniak kalibrasiemodel was in ooreenstemming met die standaard fout van die laboratorium (SEL = 1.54 mg/L). Die R2 (92.05) en die gemiddelde RVA (3.3) dui op ‘n model met uitstekende presiesheid wat gereed is vir oordra en gebruik deur die industrie. Die R2 waardes vir die alkohol-, pH- en invertsuiker –kalibrasie-modelle wat met ATR en transmissie vir die likeurmonsters ontwikkel is, dui op goeie tot uitstekende presiesheid (80<R2<100). Die alkoholkalibrasiemodel wat ontwikkel is in transmissie, is geskik vir kwaliteits- en prosesbeheerdoelwitte (RVA = 21.2), terwyl die invertsuiker kalibrasiemodel geskik is vir kwaliteitsbeheer doelwitte (5<RVA<6.4). Die pH en invertsuiker kalibrasiemodelle in ATR is geskik vir vinnige evalueringsdoelwitte, met RVA = 3.6 en RVA = 4.8 waardes, onderskeidelik. Hierdie algoritmes is ook in die kelder geimplementeer. Die pH kalibrasiemodel in transmissie was geskik vir vinnige evalueringsdoelwitte (RVA = 2.7) en toekomstige ontwikkeling is nodig om die voorspellingsakkuraatheidvan die model te verbeter. Die resultate van hierdie studie het ‘n betekenisvolle bydrae gelewer tot bevestiging van infrarooi spektroskopie as 'n kragtige tegnologie vir die vinnige kwantifisering van gehalteparameters in druiwesap en spiritus-gebaseerde likeurs. Die bydrae is veral waardevol in die konteks van voortgesette navorsing om die kwaliteit van produkte by die kelder te verbeter en aan verbruikerseise te voldoen. Die studie vir die kwantifisering van gehalteparameters in spiritus-gebaseerde likeurs is eerste in sy soort en een van die gerapporteerde verslae vir die implementering van infrarooi spektroskopie vir gehaltebeheer van Suid-Afrikaanse spiritus-gebaseerde likeurs.

Thesis (MSc)--Stellenbosch University, 2015. / ENGLISH ABSTRACT: Grape quality is linked to the organoleptic properties of grapes, raisins and wine. Many advances have been made in understanding the grape components that are important in the quality of wines and other grape products. A better understanding of the compositional content of grapes entails knowing when and how the various components accumulate in the berry. Therefore, an appreciation of grape berry development is vitally important towards the understanding of how vineyard practices can be used to improve the quality of grapes and eventually, wines. The more established methods for grape berry quality assessment are based on gravimetric methods such as colorimetry, fluorescence and chromatography. These conventional methods are accurate at targeting particular components, but are typically multi-step, destructive, expensive, polluting procedures that might be technically challenging. Very often grape berries are evaluated for quality (only) at harvest. This remains a necessary exercise as it helps viticulturists and oenologists to estimate some targeted metabolite profiles that are known to greatly influence chemical and sensory profiles of wines. However, a more objective measurement of predicting grape berry quality would involve evaluation of the grapes throughout the entire development and maturation cycle right from the early fruit to the ripe fruit. To achieve this objective, the modern grape and wine industry needs rapid, reliable, simpler and cost effective methods to profile berry development. By the turn of the last millennium, developments in infrared instrumentation such as Fourier-transform infrared (FT NIR) and attenuated total reflectance Fourier-transform infrared spectroscopy (ATR FT-IR) in combination with chemometrics resulted in the development of rapid methods for evaluating the internal and external characteristics of fresh fruit, including grapes. The advancement and application of these rapid techniques to fingerprint grape compositional traits would be useful in monitoring grape berry quality. In this project an evaluation of grape berry development was investigated in a South African vineyard setting. To achieve this goal, Sauvignon blanc grape berry samples were collected and characterised at five defined stages of development: green, pre-véraison, véraison, post-véraison and ripe. Metabolically inactivated (frozen in liquid nitrogen and stored at -80oC) and fresh berries were analysed with FT-IR spectroscopy in the near infrared (NIR) and mid-infrared (MIR) ranges to provide spectral data. The spectral data were used to provide qualitative (developmental stage) and quantitative (metabolite concentration of key primary metabolites) information of the berries. High performance liquid chromatography (HPLC) was used to separate and quantify glucose, fructose, tartaric acid, malic acid and succinic acid which provided the reference data needed for quantitative analysis of the spectra. Unsupervised and supervised multivariate analyses were sequentially performed on various data blocks obtained by spectroscopy to construct qualitative and quantitative models that were used to characterise the berries. Successful treatment of data by principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) gave statistically significant chemometric models that discriminated the berries according to their stages of development. The loadings from MIR models highlighted the important discriminant variables responsible for the observed developmental stage classification. The best calibration models to predict metabolite concentrations were obtained from MIR spectra for glucose, fructose, tartaric acid and malic acid. The results showed that both NIR and MIR spectra in combination with multivariate analysis could be reliably used to evaluate Sauvignon blanc grape berry quality throughout the fruit’s development cycle. Moreover, the methods used were fast and required minimal sample processing and no metabolite extractions with organic solvent. In addition, the individual major sugar and organic acids were accurately predicted at the five stages under investigation. This study provides further proof that IR technologies are robust and suitable to explore high-throughput and in-field application of grape compound profiling. / AFRIKAANSE OPSOMMING: Druifkwaliteit word gekoppel aan die organoleptiese eienskappe van druiwe, rosyntjies en wyn. Baie vooruitgang is reeds gemaak in die begrip van druifkomponente wat belangrik is vir die kwaliteit van wyn en ander druifprodukte. ’n Beter begrip van die samestellende inhoud van druiwe behels om te weet wanneer en hoe die verskeie komponente in die korrel opgaar. ’n Evaluasie van druiwekorrel-ontwikkeling is dus uiters belangrik vir ’n begrip van hoe wingerdpraktyke gebruik kan word om die kwaliteit van druiwe, en uiteindelik van wyne, te verbeter. Die meer gevestigde maniere vir die assessering van druiwekorrelkwaliteit is gebaseer op gravimetriese metodes soos kolorimetrie, fluoressensie en chromatografie. Hierdie konvensionele metodes is akkuraat om spesifieke komponente te teiken, maar behels tipies veelvuldige stappe en is prosesse wat destruktief en duur is, besoedeling veroorsaak, asook moontlik tegnies uitdagend is. In baie gevalle word druiwekorrels (eers) tydens oes vir kwaliteit geëvalueer. Hierdie is steeds ’n noodsaaklike oefening omdat dit wingerdkundiges en wynkundiges help om die metabolietprofiele wat daarvoor bekend is om ’n groot invloed op die chemiese en sensoriese profiele van wyn te hê en dus geteiken word, te skat. ’n Meer objektiewe meting om druiwekorrelkwaliteit te voorspel, sou die evaluering van die druiwe dwarsdeur hulle ontwikkeling- en rypwordingsiklus behels, vanaf die vroeë vrugte tot die ryp vrugte. Om hierdie doelwit te behaal, benodig die moderne druiwe- en wynbedryf vinnige, betroubare, eenvoudiger en kostedoeltreffende metodes om ’n profiel saam te stel van korrelontwikkeling. Aan die einde van die vorige millennium het ontwikkelings in infrarooi instrumentering soos Fourier-transform infrarooi (FT NIR) en attenuated total reflectance Fourier-transform infrarooi spektroskopie (ATR FT-IR) in kombinasie met chemometrika gelei tot die ontwikkeling van vinnige metodes om die interne en eksterne kenmerke van vars vrugte, insluitend druiwe, te meet. Die vooruitgang en toepassing van hierdie vinnige tegnieke om ‘vingerafdrukke’ te bekom van die samestellende kenmerke sal nuttig wees vir die verbetering van druiwekorrelkwaliteit. In hierdie projek is ’n evaluering van druiwekorrelontwikkeling in ’n Suid-Afrikaanse wingerdligging ondersoek. Ten einde hierdie doel te bereik, is Sauvignon blanc druiwekorrelmonsters op vyf gedefinieerde stadiums van ontwikkeling versamel en gekarakteriseer: groen, voor deurslaan, deurslaan, ná deurslaan en ryp. Metabolies geïnaktiveerde (bevrore in vloeibare stikstof en gestoor teen -80oC) en vars korrels is met FT-IR spektroskopie in die naby infrarooi (NIR) and mid-infrarooi (MIR) grense geanaliseer om spektrale data te verskaf. Die spektrale data is gebruik om kwalitatiewe (ontwikkelingstadium) en kwantitatiewe (metabolietkonsentrasie van belangrikste primêre metaboliete) inligting van die korrels te verskaf. High performance liquid chromatography (HPLC) is gebruik om glukose, fruktose, wynsteensuur, appelsuur en suksiensuur te skei en te kwantifiseer, wat die verwysingsdata verskaf het wat vir die kwantitatiewe analise van die spektra benodig word. Ongekontroleerde en gekontroleerde meervariantanalises is opeenvolgend op verskeie datablokke uitgevoer wat met spektroskopie verkry is om kwalitatiewe en kwantitatiewe modelle te verkry wat gebruik is om die korrels te karakteriseer. Suksesvolle behandeling van die data deur hoofkomponent analise (principal component analysis (PCA)) en ortogonale parsiële kleinstekwadraat diskriminant analise (partial least squares discriminant analysis (OPLS-DA)) het statisties betekenisvolle chemometriese modelle verskaf wat die korrels op grond van hulle ontwikkelingstadia onderskei het. Die ladings vanaf die MIR-modelle het die belangrike diskriminantveranderlikes beklemtoon wat vir die klassifikasie van die waargenome ontwikkelingstadium verantwoordelik is. Die beste kalibrasiemodelle om metabolietkonsentrasies te verkry, is vanuit die MIR-spektra vir glukose, fruktose, wynsteensuur en appelsuur bekom. Die resultate toon dat beide die NIR- en MIR-spektra, in kombinasie met meervariantanalise, betroubaar gebruik kan word om Sauvignon blanc druiwekorrelkwaliteit dwarsdeur die vrug se ontwikkelingsiklus te evalueer. Verder is die metodes wat gebruik word, vinnig en het hulle minimale monsterprosessering en geen metabolietekstraksies met organiese oplosmiddel benodig nie. Daarbenewens is die vernaamste suiker en organiese sure individueel akkuraat voorspel op die vyf stadia wat ondersoek is. Hierdie studie verskaf verdere bewys dat IR-tegnologieë robuus en geskik is om hoë-deurset en in-veld toepassings van profielsamestelling van druiweverbindings te ondersoek.

This master’s thesis studies the effect of weathering on thermally modified Norway spruce (Picea abies) and Scots pine (Pinus sylvestris) with different surface treatments. Most importantly colour changes were analysed, especially, the greying. However, other aspects of appearance changes, like cracks, mould and chemical changes on the surface were included. Special emphasis was on analysing the influence of tree species, type of thermal modification and surface treatment on these properties. Furthermore, it was tested if near infra-red (NIR) spectroscopy allows to estimate the colour, in addition, to measuring chemical changes. The whole study was set up as a decking of a gangway in Northern Sweden and evaluated after the first year of exposure. With the help of colorimetry, changes in colour based on the CIE L*C*hab colour space were measured. Test results showed that within one year all surfaces turned greyer significantly due to changes in content of lignin and cellulose measured with NIR spectroscopy. Differences could neither be observed between the uses of the two tree species nor between the uses of the thermal modifications, pressurised saturated steam at a temperature of 180 °C and superheated steam at a temperature of 212 °C. However, the surface treatment affects the colour change. Timber treated with a silicon based treatment had from the beginning a greyer colour and turned greyest after one year, while oil and pigmented oil stain slowed down the greying compared to untreated and iron vitriol treated timber. After one year of exposure for none of the treatments the colour had stabilised. Qualitative analysis of cracks and mould growth on the surface indicated some dependence on thermal modification and surface treatment. The PLS model for the prediction was not good, so no universally valid conclusions could be drawn of them. Timber with silicon based treatment showed a tendency for mould growth and timber thermally modified with pressurised saturated steam at a temperature of 180 °C tends to have cracks more often. It was possible to estimate the colour from NIR spectroscopy. Best estimations were achieved for the Chroma, followed by lightness and hue. Even better prediction of the Chroma could be achieved by fitting different models based on the surface treatments.  Hence, NIR spectroscopy allows a good estimation of the greying without needing a further measurement instrument, like a colorimeter.

Statistical depth functions became well known nonparametric tool of multivariate data analyses. The most known depth functions include the halfspace depth. Although the halfspace depth has many desirable properties, some of its properties may lead to biased and misleading results especially when data are not elliptically symmetric. The thesis introduces 2 new classes of the depth functions. Both classes generalize the halfspace depth. They keep some of its properties and since they more respect the geometric structure of data they usually lead to better results when we deal with non-elliptically symmetric, multimodal or mixed distributions. The idea presented in the thesis is based on replacing the indicator of a halfspace by more general weight function. This provides us with a continuum, especially if conic-section weight functions are used, between a local view of data (e.g. kernel density estimate) and a global view of data as is e.g. provided by the halfspace depth. The rate of localization is determined by the choice of the weight functions and theirs parameters. Properties including the uniform strong consistency of the proposed depth functions are proved in the thesis. Limit distribution is also discussed together with some other data depth related topics (regression depth, functional data depth)...

Observações ou outliers estão quase sempre presentes em qualquer conjunto de dados, seja ele grande ou pequeno. Isso pode ocorrer por erro no armazenamento dos dados ou por existirem realmente alguns pontos diferentes dos demais. A presença desses pontos pode causar distorções nos resultados de modelos e estimativas. Por isso, a sua detecção é muito importante e deve ser feita antes do início de uma análise mais profunda dos dados. Após esse diagnóstico, pode-se tomar uma decisão a respeito dos pontos atípicos. Uma possibilidade é corrigi-los caso tenha ocorrido erro na transcrição dos dados. Caso sejam pontos válidos, eles devem ser tratados de forma diferente dos demais, seja com uma ponderação, seja com uma análise especial. Nos casos univariado e bivariado, o outlier pode ser detectado analisando-se o gráfico de dispersão que mostra o comportamento de cada observação do conjunto de dados de interesse. Se houver pontos distantes da massa de dados, eles devem ser considerados atípicos. No caso multivariado, a detecção por meio de gráficos torna-se um pouco mais complexa porque a análise deveria ser feita observando-se duas variáveis por vez, o que tornaria o processo longo e pouco confiável, pois um ponto pode ser atípico com relação a algumas variáveis e não ser com relação a outras, o que faria com que o resultado ficasse mascarado. Neste trabalho, alguns métodos robustos para detecção de outliers em dados multivariados são apresentados. A aplicação de cada um dos métodos é feita para um exemplo. Além disso, os métodos são comparados de acordo com o resultado que cada um apresentar para o exemplo em questão e via simulação. / Unusual observations or outliers are frequent in any data set, if it is large or not. Outliers may occur by typing mistake or by the existence of observations that are really different from the others. The presence of this observations may distort the results of models and estimates. Therefore, their detection is very important and it is recommended to be performed before any detailed analysis, when a decision can be taken about these atypical observations. A possibility is to correct these observations if the problem occurred with the construction of the data set. If the observations are correct, different strategies can be adopted, with some weights or with special analysis. In univariate and bivariate data sets, outliers can be detected analyzing the scatter plot. Observations distant from the cloud formed by the data set are considered unusual. In multivariate data sets, the detection of outliers using graphics is more difficult because we have to analyse a couple of variables each time, which results is a long and less reliable process because we can find an observation that is unusual for one variable and not unusual for the others, masking the results. In this work, some robust methods for detection of multivariate outliers are presented. The application of each one is done for an example. Moreover, the methods are compared by the results of each one in the example and by simulation.

A presente proposta correlaciona como o investimento direto estrangeiro pode, e deve, promover o desenvolvimento sustentável no país que o acolhe. O investimento direto estrangeiro é capaz de promover uma série de vantagens competitivas quando aplicado de forma coerente como: ganhos de tecnologia, geração de empregos, capacitação de mão de obra e outros benefícios que vão além do econômico. No entanto, há a necessidade de uma mensuração para que a promoção do desenvolvimento gerado se dê de forma sustentável, para que os benefícios obtidos para a nação sejam maiores do que a degradação ambiental, emissão de poluentes e os impactos sociais causados. Utilizando um referencial sobre o investimento direto estrangeiro e desenvolvimento sustentável, o trabalho consiste em verificar se o investimento direto estrangeiro promove o desenvolvimento sustentável. Para isso, uma construção foi feita a partir de dados secundários que pudessem verificar a latência dos constructos de sustentabilidade e assim relacioná-los com o investimento direto estrangeiro com a divisão em setor primário, secundário e terciário. Com essas correlações foi possível verificar como o investimento tem impactado não apenas na economia nacional mas também qual impacto ambiental e social ele trouxe. Posteriormente, uma análise de cluster e discriminante foram feitas com o intuito de agrupar e classificar os impactos do investimento direto estrangeiro utilizando a poupança líquida ajustada, que é um indicador de sustentabilidade promovido pelo Banco Mundial. Essa construção foi possível através das técnicas de análise multivariada de dados que permitiram a relação de variáveis de diferentes categorias e se mostrou adequada para pesquisas de carácter exploratório. As evidências provenientes da discussão desse trabalho contribuem com a recente literatura que busca por estudos que relacionem o investimento direto estrangeiro e o desenvolvimento que eles promovem, melhorando assim a tomada de decisão na captura desses recursos. O trabalho contribui em verificar a possível falta de políticas públicas que integrem as dimensões de desenvolvimento sustentável. Também são apresentadas algumas variáveis que podem auxiliar na busca pelo desenvolvimento sustentável. / The present work aims to highlight the need to correlate how foreign direct investment can, and should, promote sustainable development in the country that hosts it. Foreign direct investment is able to promote a number of competitive advantages when applied consistently such as gains in technology, job creation, training of manpower and other benefits that go beyond the economic level. However, a measurement is necessary so that the promotion of development generated can be satisfactory and thus fulfill the purpose of everyone involved with this financial, environmental and social capital which was invested, so environmental degradation, emissions, incentives offered and social impacts are no greater than the benefits for the nation. By using a reference framework on foreign direct investment and sustainable development, this work aims to formulate hypotheses so this investment can be measured and be considered sustainable. In order to do this, a construction from secondary data to verify the latency of the constructs of sustainability will be made enabling to relate them to foreign direct investment in every sector nationwide. With these correlations will be possible to verify how the investment has impacted not only in the national economy but which environmental and social impact it has brought. Later, a cluster analysis and discriminant will be carried out enabling to group and classify the impacts of foreign direct investment using the adjusted net saving, which is an indicator of sustainability promoted by the World Bank. This construction will be possible through techniques of multivariate data analysis, allowing the relationship of variables of different categories, which is adequate for an exploratory research. The evidences arising from the discussion of this study contribute to the recent literature that searches for studies that relate foreign direct investment and development they promote, thereby improving decision making in the capture of these resources. The study aims to verify the possible lack of public policies that promote sustainable development dimensions. Some variables are also presented which may also contribute to the search for sustainable development.

Submitted work concerns with the theoretical and practical requirements for an automatic characterization of samples by means of Laser-Induced Breakdown Spectroscopy (LIBS). Theoretical aspects of laser-matter interaction, plasma expansion, and plasma emission are described theoretically within this work. The description of the plasma emission is enhanced with the spectral detection systems and statistical properties of the plasma. The principle of the automatic characterization is based on the multivariate data analysis theoretical background, which presents recent trend and fundamental approach for automatic spectra analysis in LIBS. Theoretical knowledge is manifested in six applications, which are presented as a comment to published manuscripts. These publications push the frontiers of automatic spectra processing in LIBS.

The scatter halfspace depth is a quite recently established concept which extends the idea of the location halfspace depth for positive definite matrices. It provides an interest- ing insight into the problem of suitability quantification of a matrix for the description of the covariance structure of the multivariate distribution. The thesis focuses on the investigation of theoretical properties of the depth for both general and more specific probability distributions which can be used for data analysis. It turns out that the es- timators of scatter parameters based on the empirical scatter depth are quite effective even under relatively weak assumptions. These estimators are useful especially for dealing with a sample containing outliers or contaminating observations. 1

In food manufacturing, the quality control procedure is a critical activity that consists in organizing, measuring, tracking, and filing the conditions of the production process and the final product, with the goal of guaranteeing the designed quality standard. During the last 30 years, due to a mounting concern by both consumers and lawmakers, the definition of quality and the application of quality control improved drastically, and new methodologies have been developed to ensure better control of food production and to understand the effect of raw materials and the process condition on the final quality of the food product. This thesis discusses the approaches to quality control procedures in food manufacture, focusing on the relationship between the conditions of the process and the quality profile of the final product, testing in a real-case scenario of a complex production process advanced data analysis procedures. The statistical and analytical procedures proposed have been applied in a real case studio from Trentingrana cheese production, a dairy consortium in the northeast region of Italy producing a ripened semi-artisanal hard cheese under the Protected Denomination of Origin (PDO) of Grana Padano. The aim is developing tailored statistical procedures that infer the effect of the critical factors of production on quality properties of this PDO product considering its semi-artisanal production process and the presence of multiple confounding factors. The statistical analyses were applied to a dataset of measurements of physical, sensory, and chemical properties collected on cheese wheels sampled systematically to represent the variability of the production of the Trentingrana wheels over two years of production. In the first introductory chapter, after a review of the different definitions of quality, the most important quality parameters for a food product and the standard measurement techniques adopted in quality control are presented. Then, in the chapter 2, the standard procedures of data analysis are reviewed, as well as the new approaches derived from the context of the foodomic sciences and machine learning models for the analysis of quality control data in food manufacturing. Two implemented and tested practical statistical procedures in the context of the Trentingrana consortium are reported: the results are discussed according to the objectives of the quality control process, the type of data, and the organization of food production. In the first case, reported in chapter 3, Linear Mixed Model ANOVA Simultaneous Component Analysis (LMM-ASCA) was developed to investigate the effect of the dairy factory, the bimester of production, and the variability within a cheese wheel using colorimetric and textural measurements. In the second case, reported in chapter 4, a standard ASCA model with the addition of a blocking factor to include systematic error was developed to investigate the relationship between the dairy factory and bimester of production and the volatile organic compounds (VOCs) profile of Trentingrana cheese wheels. In addition, in chapter 5, an approach to relate physical measurements on Trentingrana samples with sensory evaluations of texture by a trained panel is presented. The objective of this procedure is to incorporate the quality control procedure information from different quality parameters. The development of the Partial Least Squares (PLS) predictive model, its validation, and the evaluation of its performances are discussed. In the last section (chapter 6), the development of an image analysis procedure to measure the visual quality of the rind thickness of cheese wheels is reported, comparing the performances of two different algorithms. The data analysis tools proposed in this thesis have been proved to be useful for exploring, inferring, and plotting the process quality properties and suitable for analyzing complex and unbalanced experimental designs. Furthermore, the data analysis procedures proposed improve quality control activity both at the process level and at the product level, increasing the information that is possible to extract from the measurement collected in a context where standard statistical approaches cannot infer significant information.