<p>The Sub-Saharan Africa (SSA) region is faced with limited medical personnel and healthcare services to address the many healthcare problems of the region. Poor health indicators reflect the overall decline in socio-economic development. Shortages of access to health services in the region is further complicated by the concentration of health services in urban areas, the region’s multiple medical problems (over 70% of HIV/AIDS cases in the world); and the brain drain phenomenon – it is estimated one-third of African physicians emigrate to North America and Europe. The result is that the SSA region is left with about 10 physicians, and 20 beds, per 100,000 patients. Telemedicine has been found to offer socio-economic benefits, reduce costs, and improve access to healthcare service providers by patients, but previous attempts to move various information technologies from developers in the industrial world to the developing world have failed because of a clear neglect of infrastructural and cultural factors that influence such transfers. The objective of this study is to address key factors that challenge the introduction of telemedicine technology into the health sector in SSA in particular, and by extension, other developing countries with similar socio-economic structures.</p><p>This research offers a distinctive perspective, focusing on visually-based clinical applications in the SSA region, and considerable attention to the national infrastructure and cultural impact of telemedicine transfer (social and value) outcomes. Two research models and its associated hypotheses are proposed and empirically tested using quantitative data collected from SSA physicians and other health professionals. The study also contributes to the ongoing debate on the potential of telemedicine in improving access and reducing costs. This research can help to understand the socio-economic impact of telemedicine outcomes in a comprehensive way. The finding from the survey shows the rapid advances in telemedicine technology specifically, visual clinical applications may become an essential healthcare tool in the near future within SSA countries.</p>

Lean management is becoming the standard for systematic productivity improvement, but the majority of implementations fail to sustain. Hence, the critical success factors for lean were the focus of this work. Literature review showed that the causality for lean success was not empirically developed beyond case study contextualisation. A multifaceted work was developed with contextualisation studies, survey of lean knowledge (758 responses), and a comprehensive case-study questionnaire (1253 responses from 44 countries). The statistical methods included exploratory factor analysis and path analysis by structural equation modelling (SEM). The first questionnaire revealed two different understandings of lean, and the second explored the underlying causality for lean success, including contingency for business size and product variety. Many contributions to the body of knowledge issued from this work. First of all, there was a methodological contribution, pioneering explorative structural modelling of full scope lean implementation. Second, SEMs of the lean knowledge-based view showed the profound positive effects of management knowledge on the primary factors for lean success. These factors were shown to be leadership and employee development. Third, the most beneficial lean methods were highlighted for specific scenarios. Fourth, the negligible and negative effects of a consultant-based approach to lean were uncovered. The results showed that the majority of consultants did not aid the long-term performance and sustainability of lean but significantly hindered it, except where masterful consultants acted as coaches. Fifth, a shortage of lean knowledge was observed in New Zealand; their participants averaged only half of what the USA�s did. Sixth, as culture has been emphasised in current literature, the present danger of overly focusing on it was discussed. Seventh was a conceptual contribution integrating lean and risk management, and a practical application with a risk analysis. This developed a risk matrix for the assessment and prioritisation of implementation components. Eighth, some adjustments to government lean strategies were proposed. And finally, the work integrated the findings in a tangible stage process model for implementation in SMEs. The dissemination of this knowledge has the potential to enhance productivity and commercial success of industries in New Zealand and abroad through successful lean implementations. Lean is not a weak methodology but it has been misunderstood and misapplied.

Este trabalho apresenta um estudo comparativo do poder de predição de quatro métodos de regressão adequados para situações nas quais os dados, dispostos na matriz de planejamento, apresentam sérios problemas de multicolinearidade e/ou de alta dimensionalidade, em que o número de covariáveis é maior do que o número de observações. No presente trabalho, os métodos abordados são: regressão por componentes principais, regressão por mínimos quadrados parciais, regressão ridge e LASSO. O trabalho engloba simulações, em que o poder preditivo de cada uma das técnicas é avaliado para diferentes cenários definidos por número de covariáveis, tamanho de amostra e quantidade e intensidade de coeficientes (efeitos) significativos, destacando as principais diferenças entre os métodos e possibilitando a criação de um guia para que o usuário possa escolher qual metodologia usar com base em algum conhecimento prévio que o mesmo possa ter. Uma aplicação em dados reais (não simulados) também é abordada. / This paper presents a comparative study of the predictive power of four suitable regression methods for situations in which data, arranged in the planning matrix, are very poorly multicolinearity and / or highdimensionality, wherein the number of covariatesis greater the number of observations. In this study, the methods discussed are: principal component regression,partial least squares regression,ridge regression and LASSO. The work includes simulations, where in the predictive power of each of the techniques is evaluated for different scenarios defined by the number of covariates, sample size and quantity and intensity ratios (effects) significant, high lighting the main dffierences between the methods and allowing for the creating a guide for the user to choose which method to use based on some prior knowledge that it may have. An applicationon real data (not simulated) is also addressed.

[pt] Muitos problemas da área de aprendizagem automática tem por objetivo modelar a complexa relação existente num sisitema , entre variáveis de entrada X e de saída Y na ausência de um modelo teórico. A regressão por mínimos quadrados parciais PLS ( Partial Least Squares) constitui um método linear para resolução deste tipo de problema , voltado para o caso de um grande número de variáveis de entrada quando comparado com número de amostras. Nesta tese , apresentamos uma variante do algoritmo clássico PLS para o tratamento de grandes conjuntos de dados , mantendo um bom poder preditivo. Dentre os principais resultados destacamos um versão paralela PPLS (Parallel PLS ) exata para o caso de apenas um variável de saída e um versão rápida e aproximada DPLS (DIRECT PLS) para o caso de mais de uma variável de saída. Por outro lado ,apresentamos também variantes para o aumento da qualidade de predição graças à formulação não linear. São elas o LPLS ( Lifted PLS ), algoritmo para o caso de apenas uma variável de saída, baseado na teoria de funções de núcleo ( kernel functions ), uma formulação kernel para o DPLS e um algoritmo multi-kernel MKPLS capaz de uma modelagemmais compacta e maior poder preditivo, graças ao uso de vários núcleos na geração do modelo. / [en] The purpose of many problems in the machine learning field isto model the complex relationship in a system between the input X and output Y variables when no theoretical model is available. The Partial Least Squares (PLS)is one linear method for this kind of problem, for the case of many input variables when compared to the number of samples. In this thesis we present versions of the classical PLS algorithm designed for large data sets while keeping a good predictive power. Among the main results we highlight PPLS (Parallel PLS), a parallel version for the case of only one output variable, and DPLS ( Direct PLS), a fast and approximate version, for the case fo more than one output variable. On the other hand, we also present some variants of the regression algorithm that can enhance the predictive quality based on a non -linear formulation. We indroduce LPLS (Lifted PLS), for the case of only one dependent variable based on the theory of kernel functions, KDPLS, a non-linear formulation for DPLS, and MKPLS, a multi-kernel algorithm that can result in a more compact model and a better prediction quality, thankas to the use of several kernels for the model bulding.

This project was dedicated to the problem of improving a non-invasive blood glucose monitor being developed by the VivaScan Corporation. The company has made some progress in the non-invasive blood glucose device development and approached WPI for a statistical assistance in the improvement of their model in order to predict the glucose level more accurately. The main goal of this project was to improve the ability of the non-invasive blood glucose monitor to predict the glucose values more precisely. The goal was achieved by finding and implementing the best regression model. The methods included ordinary least squared regression, partial least squares regression, robust regression method, weighted least squares regression, local regression, and ridge regression. VivaScan calibration data for seven patients were analyzed in this project. For each of these patients, the individual regression models were built and compared based on the two factors that evaluate the model prediction ability. It was determined that partial least squares and ridge regressions are two best methods among the others that were considered in this work. Using these two methods gave better glucose prediction. The additional problem of data reduction to minimize the data collection time was also considered in this work.

The mobile banking is an increasingly popular service for customers of the traditional banking industry. On the surface, China has the highest adoption rate of this new technology, yet many users do not remain active or they only use mobile banking for the simplest tasks such as checking their balance. This research was designed to uncover the reasons for these two issues by identifying the major factors influencing users' intention to continue using mobile banking (continuance intention) as well as their behavioural intention to try new mobile banking functions. To do so, an integrated model was developed on the basis of the unified theory of acceptance and use of technology model (UTAUT), task-technology fit model (TTF) and expectation confirmation model (ECM). Empirical data were collected from China's mobile banking users and the integrated model was tested using Partial Least Squares Structural Equation Modelling (PLS-SEM) approach. The results indicate that the main factors which positively influence the continuance intention include satisfaction, performance expectancy and effort expectancy. Continuance intention, in turn, influences behavioural intention to try new functionality, together with social influence, facilitating conditions and confirmation. Moreover, some mediating effects were discovered. For example, task-technology fit may indirectly affect the continuance intention through users' satisfaction. The research results have a number of theoretical contributions. Firstly, this research discovers that the impact of task-technology fit on users' continuance intention towards mobile banking is fully mediated by users' satisfaction. This enriches the extant literature that is mostly focused on the technology perceptions (e.g. performance expectancy and effort expectancy) of users. Secondly, this research identifies that satisfaction also mediates the impact of confirmation of expectations on continuance intention, also extending the literature on the continuance usage of information systems. Thirdly, this research fills the gap in extant research regarding users' intention to try new mobile banking functions, by proposing a new integrated model using constructs from UTAUT, TTF, and ECM, and demonstrating that continuance intention itself fully mediates the impact of performance expectancy and effort expectancy on behavioural intention. The new model has a high explanatory power than each individual model offers. The research results also have management implications in terms of how to improve the task-technology fit to support continuous use and extended the use of mobile banking. For instance, to improve continuance intention banks can improve satisfaction by optimising task-technology fit. This in turn will require better understanding of users' different task requirements in specific market segments. In addition, banks can also attract users' behavioural intention to try new functionality by timely updating corresponding technology and launching marketing campaigns to keep users informed of any new functions of their mobile banking application.

This project utilised a novel solvent-free continuous cocrystallisation (SFCC) method to manufacture pharmaceutical co-crystals. The objectives were to optimize the process towards achieving high co-crystal yields and to understand the behaviour of co-crystals under different conditions. Particular attention was paid to the development of near infrared (NIR) spectroscopy as a process analytical technology (PAT). Twin screw, hot melt extrusion was the base technique of the SFCC process. Changing parameters such as temperature, screw speed and screw geometry was important for improving the co-crystal yield. The level of mixing and shear was directly influenced by the screw geometry, whilst the screw speed was an important parameter for controlling the residence time of the material during hot melt extrusion. Ibuprofen – nicotinamide 1:1 cocrystals and carbamazepine – nicotinamide 1:1 co-crystals were successfully manufactured using the SFCC method. Characterisation techniques were important for this project, and NIR spectroscopy proved to be a convenient, accurate analytical technique for identifying the formation of co-crystals along the extruder barrel. Separate thermal and model shear deformation studies were also carried out to determine the effect of temperature and shear on co-crystal formation for several different pharmaceutical co-crystal pairs. Finally, NIR spectroscopy was used to create two partial least squares regression models, for predicting the 1:1 co-crystal yield of ibuprofen – nicotinamide and carbamazepine – nicotinamide, when in a powder mixture with the respective pure API. It is believed that the prediction models created in this project can be used to facilitate future in-line PAT studies of pharmaceutical co-crystals during different manufacturing processes.

Os sistemas de análises em fluxo com multi-impulsão (MPFS) têm como característica principal o emprego de bombas solenóide como unidade propulsora de fluidos, as quais proporcionam fluxo pulsado. Este regime de fluxo foi avaliado em função das condições de mistura entre as soluções envolvidas, transferência de calor e difusão gasosa. A associação dos métodos quimiométricos de análises e dos sistemas MPFS foi demonstrada em relação à determinação espectrofotométrica de glicose, frutose e glicerol em vinhos fermentiscíveis e caldos de cana-de-açúcar. O método se fundamentou na reação dos carboidratos com metaperiodato de sódio e posterior oxidação de iodeto pelo metaperiodato remanescente com monitoramento de [I3 -] produzido. O tratamento dos dados envolveu calibração multivariada, empregando o algoritmo PLS e os resultados são concordantes com aqueles obtidos por cromatografia líquida de troca aniônica com detecção por amperometria pulsada. O sistema proposto é simples e robusto, capaz de analisar 120 amostras por hora. O fluxo pulsado proporcionou melhoria no desenvolvimento reacional no que diz respeito à transferência de calor e difusão gasosa. Esse aspecto se deve principalmente ao aumento do transporte de massas no sentido radial. Estes fatos foram constatados na determinação espectrofotométrica de açúcares redutores totais (ART) e etanol. O sistema MPFS proposto para a determinação de ART envolveu hidrólise ácida da sacarose e degradação alcalina dos carboidratos. A natureza do fluxo pulsado possibilitou o uso de menores temperaturas de um banho termostatizado durante as etapas de hidrólise e degradação, bem como a diminuição da alcalinidade. Para a mecanização da determinação espectrofotométrica de etanol envolvendo a redução de Cr(VI) a Cr(III) sob condições ácidas, foi desenvolvido um sistema MPFS, o qual se demonstrou eficiente e adequado para procedimentos que envolvem difusão gasosa. Após otimização dos principais parâmetros envolvidos, os mesmos foram comparados empregando o sistema de multi-comutação, cujo fluxo é laminar. Melhores resultados analíticos foram obtidos no sistema proposto, que resultou em boa sensibilidade. Em relação aos sistemas de análises em fluxo que exploram interface reacional única (SIFA), foram demonstradas suas potencialidades através da implementação de procedimentos que envolvem determinações simultâneas, sem a necessidade de reconfigurações no módulo de análises. Ainda, a simplificação da etapa de otimização foi espectrofotometricamente avaliada através da determinação de alumínio, ferro total e P-PO4. O sistema proposto é de configuração simples e capaz de analisar 130, 140 e 90 amostras de alumínio, ferro total e fósforo por hora, respectivamente / Multi-pumping flow systems (MPFS) present as an unique feature the use of solenoid pumps as fluid propelling devices, which deliver pulsed flows. This flow regime was evaluated in order to improve mixing conditions between the involved solutions, heating transfer and gas diffusion.The association of the chemometric methods of analysis and MPFS systems was demonstrated in the spectrophotometric determination of glucose, fructose and glycerol in musts and sugar cane juices. The method involved metaperiodate oxidation of carbohydrates and further oxidation of remainder metaperiodate iodide yield in the [I3 -] complex that was spectrophotometrically monitored. Data treatment involved multivariate calibration relying on the PLS algorithm and results were in agreement with liquid anion chromatography with pulsed amperometric detection. The proposed system is simple and rugged, allowing 120 samples to be run per hour. The pulsed flow led to a enhanced in heating transfer and gas diffusion, in view of the enhanced radial mass transport. These aspects were verified in the spectrophotometric determination of total reducing sugars (TRS) and ethanol. The proposed MPFS system for TRS determination involved in-line hydrolysis of sucrose and alkaline degradation of the carbohydrates. The intrinsic characteristic of pulsed flow allowed the use of lower temperatures in bath thermostatization during hydrolysis and degradation steps, as well as a lower alkalinity. The MPFS for spectrophotometric determination of ethanol involving diffusion towards an acceptor stream, reduction of Cr(VI) to Cr(III) under acidic condition, and Cr*(III) monitoring proved to be eficient and amenale to analytical procedures involving gas diffusion. After optimization of the main parameters, the system was compared with a multicommuted flow system (MCFA) that exploits a laminar flow. Better analytical results were obtained with the proposed system which demonstrated fair sensitivity. Regarding flow systems exploiting a single reaction interface (SIFA), their potentialities were demonstrated by implementing analytical procedures for simultaneous determination without requiring reconfigurations in the flow manifold. In this proposed system the simplification of the optimization step was atained, and the approach was evaluated in relation to spectrophotometrically determination of aluminum, total iron and phosphate. The system exhibits simple configuration and allows 130, 140 and 90 samples of aluminum, total iron and phosphate to be run per hour, respectivelly

A laranja é uma das frutas mais produzidas e consumidas no mundo, sendo o Brasil o maior produtor e exportador do seu suco concentrado. Entretanto, pragas e doenças comprometem consideravelmente sua produção. Atualmente, a doença mais preocupante é o Greening, também conhecida mundialmente como Huanglongbing (HLB). A doença não possui cura, apresenta longa fase assintomática e não possui um método eficiente de controle. Além disso, não existem métodos de diagnóstico aplicáveis em larga escala. Neste trabalho são propostas as técnicas fotônicas de fluorescência induzida por laser e de infravermelho por transformada de Fourier para o diagnóstico do HLB. Para a realização das medidas, foram coletadas folhas de árvores saudáveis, doentes com HLB e doentes com a clorose variegada dos citros, sendo esta incluída nos estudos para verificar a capacidade de diferenciação entre as doenças. Foram coletadas quatro classes de folhas nessas plantas: sadia, HLB-sintomática, HLB-assintomática e CVC-sintomática. As folhas foram medidas em laboratório e seus espectros foram pré-processados para indução de um classificador via regressão por mínimos quadrados parciais. Além das folhas, foram medidas amostras dos seguintes metabólitos primários e secundários para entendimento espectral: amido, glicose, sacarose, hesperidina, naringina e umbeliferona. Taxas de acerto de superiores a 89% foram obtidas na classificação das folhas nas técnicas de fluorescência e infravermelho, sendo superior às taxas dos métodos de manejo empregados atualmente no campo. A fluorescência induzida por laser apresenta um grande potencial para uso em campo devido a possibilidade de miniaturização de seus componentes. Os espectros dos metabólitos secundários apresentam fortes indícios de que a alteração de suas concentrações podem contribuir na detecção de doenças pelas técnicas fotônicas. / Sweet orange is one of the most produced and consumed fruit in the world, and Brazil is the largest producer and exporter of this fruit. However, pests and diseases significantly reduce the worldwide production. Currently, the most destructive disease in the field is called greening, also known as huanglongbing (HLB). There is no control for HLB. In addition, the disease presents a long asymptomatic phase. Furthermore, no diagnostic methods are available to use in large scale. In this study are proposed fluorescence and infrared spectroscopy for the HLB diagnosis. For the measurements were collected leaves from healthy, HLB- and citrus variegated chlorosis-infected plants, being the last one to comparison between the diseases. It were collected four classes of leaves: healthy, HLB-asymptomatic, HLB-symptomatic and CVC-symptomatic. The leaves were measured and their spectra were pre-processed for the induction of classifier via partial least squares regression. In addition, samples of plant metabolites were measured for leaves spectral interpretation: starch, glucose, sucrose, hesperidin, naringin and umbelliferone. Success rates above 89% were obtained through both photonic techniques, higher compared to the sucess rates of the actual management methods. The metabolites spectra have shown strong evidence that their concentrations changes could contribute to the diagnosis of the diseases by photonic techniques. Particularly, the fluorescence spectroscopy seems interesting because it has a great potential for field application due to the existence of portable photonic devices.

Submitted by admin tede (tede@pucgoias.edu.br) on 2018-05-28T19:09:38Z No. of bitstreams: 1 LEONARDO RODRIGUES DE OLIVEIRA MERELLES.pdf: 1239497 bytes, checksum: cf674b2af1fe982136ca8ff2e1740f37 (MD5) / Made available in DSpace on 2018-05-28T19:09:38Z (GMT). No. of bitstreams: 1 LEONARDO RODRIGUES DE OLIVEIRA MERELLES.pdf: 1239497 bytes, checksum: cf674b2af1fe982136ca8ff2e1740f37 (MD5) Previous issue date: 2018-03-27 / Understanding how climate change influences crop yield contributes to the forecasting of its consequences and assists in the management of agribusiness and food security. In this study, partial least squares regression (PLSR) was used to quantify the contribution of climate change to soybean and corn yield. To verify the accuracy of the model, the time series regression was applied. The Central-West and South regions of Brazil were considered for the modeling of grain yield using precipitation, temperature, humidity, evapotranspiration and carbon dioxide levels between 1980 and 2016 as predictors. In both regions, the main contributor to increased grain yield was the increase in carbon dioxide levels, and the main contributor to the reduction of productivity was air humidity, suggesting that climate change is influencing grain yield. The results confirm climate change, indicating that farmers and decision makers should work on other variables that control productivity, such as genetic improvement and agricultural management practices. / Compreender como as mudanças climáticas influenciam na produtividade agrícola contribui com a previsibilidade de suas consequências e auxilia na gestão do agronegócio e da segurança alimentar. Neste estudo, foi utilizada a regressão por mínimos quadrados parciais (PLSR) para quantificar as contribuições das mudanças climáticas na produtividade da soja e do milho. Para verificar a qualidade de ajuste do modelo foi aplicada a regressão de séries temporais. As regiões Centro-Oeste e Sul do Brasil foram consideradas para a condução da modelagem da produtividade de grãos, utilizando como variáveis preditoras a precipitação, temperatura, umidade, evapotranspiração e os níveis de dióxido de carbono, entre os anos de 1980 a 2016. Nas duas regiões, o principal contribuinte para o incremento da produtividade de grãos foi o aumento nos níveis de dióxido de carbono e o principal contribuinte para a redução da produtividade foi a umidade do ar, sugerindo que as mudanças climáticas estão influenciando na produtividade de grãos. Os resultados confirmam as mudanças climáticas, indicando que agricultores e tomadores de decisões deverão atuar em outras variáveis que controlam a produtividade, como melhoramento genético e práticas de gestão agrícola.