Multivariate control charts are valuable tools for multivariate statistical process control (MSPC) used to monitor industrial processes and to detect abnormal process behavior. It has been shown in the literature that Bayesian control charts are optimal tools to control the process compared with the non-Bayesian charts. To use any control chart, three control chart parameters must be specified, namely the sample size, the sampling interval and the control limit. Traditionally, control chart design is based on its statistical performance. Recently, industrial practitioners and academic researchers have increasingly recognized the cost benefits obtained by applying the economically designed control charts to quality control, equipment condition monitoring, and maintenance decision-making. The primary objective of this research is to design multivariate Bayesian control charts (MVBCH) both for quality control and conditional-based maintenance (CBM) applications. Although considerable research has been done to develop MSPC tools under the assumption that the observations are independent, little attention has been given to the development of MSPC tools for monitoring multivariate autocorrelated processes. In this research, we compare the performance of the squared predication error (SPE) chart using a vector autoregressive moving average with exogenous variables (VARMAX) model and a partial least squares (PLS) model for a multivariate autocorrelated process. The study shows that the use of SPE control charts based on the VARMAX model allows rapid detection of process disturbances while reducing false alarms. Next, the economic and economic-statistical design of a MVBCH for quality control considering the control limit policy proved to be optimal by Makis(2007) is developed. The computational results illustrate that the MVBCH performs considerably better than the MEWMA chart, especially for smaller mean shifts. Sensitivity analyses further explore the impact of the misspecified out-of-control mean on the actual average cost. Finally, design of a MVBCH for CBM applications is considered using the same control limit policy structure and including an observable failure state. Optimization models for the economic and economic statistical design of the MVBCH for a 3 state CBM model are developed and comparison results show that the MVBCH performs better than recently developed CBM Chi-square chart.

Multivariate control charts are valuable tools for multivariate statistical process control (MSPC) used to monitor industrial processes and to detect abnormal process behavior. It has been shown in the literature that Bayesian control charts are optimal tools to control the process compared with the non-Bayesian charts. To use any control chart, three control chart parameters must be specified, namely the sample size, the sampling interval and the control limit. Traditionally, control chart design is based on its statistical performance. Recently, industrial practitioners and academic researchers have increasingly recognized the cost benefits obtained by applying the economically designed control charts to quality control, equipment condition monitoring, and maintenance decision-making. The primary objective of this research is to design multivariate Bayesian control charts (MVBCH) both for quality control and conditional-based maintenance (CBM) applications. Although considerable research has been done to develop MSPC tools under the assumption that the observations are independent, little attention has been given to the development of MSPC tools for monitoring multivariate autocorrelated processes. In this research, we compare the performance of the squared predication error (SPE) chart using a vector autoregressive moving average with exogenous variables (VARMAX) model and a partial least squares (PLS) model for a multivariate autocorrelated process. The study shows that the use of SPE control charts based on the VARMAX model allows rapid detection of process disturbances while reducing false alarms. Next, the economic and economic-statistical design of a MVBCH for quality control considering the control limit policy proved to be optimal by Makis(2007) is developed. The computational results illustrate that the MVBCH performs considerably better than the MEWMA chart, especially for smaller mean shifts. Sensitivity analyses further explore the impact of the misspecified out-of-control mean on the actual average cost. Finally, design of a MVBCH for CBM applications is considered using the same control limit policy structure and including an observable failure state. Optimization models for the economic and economic statistical design of the MVBCH for a 3 state CBM model are developed and comparison results show that the MVBCH performs better than recently developed CBM Chi-square chart.

A big challenge in managing large scale industry processes, like the ones in the paper and pulp industry, is to reduce the amount of downtime and reduce sources of product quality variability to a minimum, while staying cost effective. To accomplish this the key is to understand the complex nature of the processes variables, and to quantify the causal relationships between them and the product quality together with the amount of output. Paper and pulp industry processes consist mainly of chemical processes and the relatively low cost of sensors today enables collection of huge amounts of data, both variables and observations on frequent time intervals. These masses of data usually come with the intrinsic problem of multicollinearity which requires efficient multivari- ate statistical tools for the extraction of useful insights among the noise. One goal in this multivariate situation is to breakthrough the noise and find a relatively small subset of variables that are important, that is, variable selection. The purpose with this master thesis is to help SCA Obbola, a large paper manu- facturer that have had a variable production output, to come up with conclusions that can help them ensure a long term high production quantity and quality. We apply different variable selection approaches that have proven successful in the literature. The results that we get are of mixed success, but we manage to find both variables that SCA Obbola knows affect specific response variables, but also variables that they find interesting for further investigation. / En stor utmaning när det gäller att hantera storskaliga industriprocesser, som i pappers- och massaindustrin, är att minska tiden för driftstopp och reducera källor till varia- tioner i produktkvalitén till ett minimum, och samtidigt vara kostnadseffektiv. För att uppnå detta är det viktigt att förstå processvariablernas komplexa natur och att kvantifiera orsakssambanden mellan dem och produktkvaliteten tillsammans med pro- duktionsmängden. Pappers- och massasindustrin består huvudsakligen av kemiska pro- cesser och den relativt låga kostnaden för sensorer idag möjliggör insamling av stora mängder data, både variabler och observationer inom frekventa tidsintervall. Med des- sa datamängder får man ofta problem med multikollinearitet, vilket kräver effektiva multivariata statistiska verktyg för att extrahera användbara insikter bland bruset. Ett mål i denna multivariata situation är att bryta igenom bruset och hitta en relativt liten delmängd variabler som är viktiga, det vill säga variabel selektion. Syftet med denna masteruppsats är att hjälpa SCA Obbola, en stor pappersprodu- cent som har haft ett varierat produktionsutfall, att komma fram till slutsatser som kan hjälpa dem att säkerställa en långsiktig hög produktionskvantitet och kvalitet. Vi tillämpar olika metoder för variabel selektion, som har visat sig framgångsrika i lit- teraturen. Resultaten av arbetet är av blandad framgång, men vi lyckas hitta både variabler som SCA Obbola vet påverkar specifika responser, men även variabler som de tycker är intressanta för vidare utredning.

abstract: In healthcare facilities, health information systems (HISs) are used to serve different purposes. The radiology department adopts multiple HISs in managing their operations and patient care. In general, the HISs that touch radiology fall into two categories: tracking HISs and archive HISs. Electronic Health Records (EHR) is a typical tracking HIS, which tracks the care each patient receives at multiple encounters and facilities. Archive HISs are typically specialized databases to store large-size data collected as part of the patient care. A typical example of an archive HIS is the Picture Archive and Communication System (PACS), which provides economical storage and convenient access to diagnostic images from multiple modalities. How to integrate such HISs and best utilize their data remains a challenging problem due to the disparity of HISs as well as high-dimensionality and heterogeneity of the data. My PhD dissertation research includes three inter-connected and integrated topics and focuses on designing integrated HISs and further developing statistical models and machine learning algorithms for process and patient care improvement. Topic 1: Design of super-HIS and tracking of quality of care (QoC). My research developed an information technology that integrates multiple HISs in radiology, and proposed QoC metrics defined upon the data that measure various dimensions of care. The DDD assisted the clinical practices and enabled an effective intervention for reducing lengthy radiologist turnaround times for patients. Topic 2: Monitoring and change detection of QoC data streams for process improvement. With the super-HIS in place, high-dimensional data streams of QoC metrics are generated. I developed a statistical model for monitoring high- dimensional data streams that integrated Singular Vector Decomposition (SVD) and process control. The algorithm was applied to QoC metrics data, and additionally extended to another application of monitoring traffic data in communication networks. Topic 3: Deep transfer learning of archive HIS data for computer-aided diagnosis (CAD). The novelty of the CAD system is the development of a deep transfer learning algorithm that combines the ideas of transfer learning and multi- modality image integration under the deep learning framework. Our system achieved high accuracy in breast cancer diagnosis compared with conventional machine learning algorithms. / Dissertation/Thesis / Doctoral Dissertation Industrial Engineering 2018

Hybrit development AB är ett bolag som strävar mot att kunna producera fossilfritt stål genom att reducera järnmalmspellets med hjälp av vätgas. Därför har Hybrit utfört experimentella kampanjer där genomförbarheten av att reducera järnmalmspellets med hjälp av vätgas undersökts och studerats. Vid produktion av järn och stål måste produktkvalitén tas i beaktan. Reduktionsprocessen karaktäriseras av en mängd olika process- och kvalitetsparametrar, där kvalitetsparametrarna beskriver produktkvalitén. Det är av intresse att studera hur processparametrarna påverkar produktkvalitén. Processparametrarna kan mätas vid vilket tidpunkt som helst genom olika sensorer. Produktkvalitén kan bestämmas först efter att järnmalmspelletsen är färdigreducerad. Därför präglas processen av en tidsfördröjning mellan mätningen av processparametrarna och labanalysemätningarna av kvalitetsparametrarna. På grund av tidsfördröjningen är det av intresse att kunna prediktera produktkvalitén utifrån processparametrarna. Om det går att prediktera produktkvalitén, är det av vikt att kunna avgöra prediktionens giltighet.  Examensarbetets syfte är att identifiera hur reduktionsprocessparametrarna påverkar reducerade järnets kvalitetsparametrar. En processövervakningsmetod som passar för processövervakning ska testas och undersökas utifrån hur metoden kan användas för att avgöra prediktionens giltighet. Processövervakningen ska användas för att avgöra om processen befinner sig i ett processläge som bidrar till en någorlunda korrekt och lämplig prediktion av produktkvalitén.  För analys av data användes 65 processparametrar och 6 kvalitetsparametrar. Den multivariata analysmetoden Partial Least Squares (PLS) användes för att nå syftet med examensarbetet. Via PLS skapades en modell som kunde beskriva vilka processparametrar som påverkade kvalitetsparametrarna samt hur processparametrarna påverkade kvalitetsparametrarna. PLS-modellen kunde prediktera kvalitetsparametrarna någorlunda korrekt och lämpligt, givet att processen befinner sig inom ramen för modellen och att det är en hög förklaringsgrad för kvalitetsparametern som predikteras. Kvalitetsparametern Y6-1 predikterades sämre eftersom förklaringsgraden för Y6-1 var låg. Processövervakningsmetoden som testades och undersöktes var PLS-övervakning. För att undersöka hur PLS-övervakning kan användas för att avgöra prediktions giltighet, användes tre processövervakningsverktyg. Dessa var X-scores processövervakning, Hotelling T2 och SPE. Resultatet var att PLS-övervakning kunde angiva hur processen förhåller sig till modellen. Observationerna som avvek i PLS-övervakningen predikterades sämre. Därmed kunde information om prediktionens giltighet genom PLS-övervakning erhållas. Att tillämpa PLS-övervakning för att avgöra prediktionens giltighet är en större framgång. Detta på grund av att information om produktkvalitén innan reduktionsprocessen är genomförd kan användas för att säkerställa produktion med tillfredställande kvalitet. Att tillämpa multivariata processövervakningsmetoder för att övervaka de predikterade kvalitetsparametrarna kan vara av intresse för framtida studier. Detta då processövervakningen kan användas för att minimera den interna variationen hos kvalitetsparametrarna. / Hybrit development AB strives to produce fossil-free steel by using hydrogen for the direct reduction process of iron ore pellets. To achieve that goal, Hybrit has carried out experimental campaigns where the feasibility of direct reduction using hydrogen gas has been investigated and studied. The quality of the reduced iron must be considered when producing iron and steel. The reduction process is characterized by a variety of process- and quality parameters. Because the quality parameters describe the quality of the product, it is of interest to study how the process parameters affect the quality parameters. The process parameters can be measured at any time through various sensors around the reactor in which the iron ore pellets are reduced. While the quality of the product can only be determined after the iron ore pellets have been completely reduced. Therefore, the process is characterized by a time delay between the measurement of the process parameters and the measurement of the quality parameters, where the reduced iron must be analyzed in a laboratory before the quality parameters can be measured. Because of the time delay, it is of interest to be able to predict the quality of the product based on the process parameters. If it is possible to predict the quality, then it is of importance to be able to determine the validity of the prediction.  The aim of this master thesis is to identify how the reduction process parameters affect the quality parameters of the reduced iron. A process monitoring method suitable for monitoring the process need be tested and investigated based on how the method can be used to determine the validity of the prediction. The process monitoring will be used to determine whether the process is in a process state that contributes to a reasonably accurate and appropriate prediction of the quality of the product.  65 process parameters and 6 quality parameters were used for the analysis of how the reduction process parameters affect the quality parameters of the reduced iron. The multivariate analysis method Partial Least Squares (PLS) was used to achieve the aim of the thesis. A multivariate model which could describe how the process parameters affect the quality parameters was created through PLS. The PLS-model was able to predict the quality parameters reasonably correctly and appropriately, given that the process is within the scope of the model and that the explanatory power is high for the quality parameter that is predicted. The quality parameter Y6-1 could not be predicted reasonably correct as the explanatory power for Y6-1 was low. The process monitoring method tested and investigated was PLS monitoring. Three process monitoring tools were used when PLS monitoring was investigated based on how they can be used to determine the validity of the prediction. These tools were X-scores process monitoring, Hotelling T2 and SPE. The result was that PLS monitoring could indicate how the process relates to the model. Observations that deviated in the PLS monitoring could not be predicted correctly. Thus, information about the validity of the prediction through PLS monitoring could be obtained. Applying PLS monitoring to determine the validity of the prediction is a greater success. This is because information about the quality of the product before the reduction process is completed can be used to ensure production with a satisfactory product quality. Applying multivariate process monitoring methods to monitor the predicted quality parameters may be of interest for future studies. This is because the process monitoring can be used to minimize the internal variation of the quality parameters.

Esta tese apresenta uma sistemática para seleção de variáveis de processo e determinação da condição ótima de operação em processos contínuos multivariados e em múltiplos estágios. O método proposto é composto por seis etapas. Um pré-tratamento nos dados é realizado após a identificação das variáveis de processo e do estabelecimento dos estágios de produção, onde são descartadas observações com valores espúrios e dados remanescentes são padronizados. Em seguida, cada estágio é modelado através de uma regressão Partial Least Squares (PLS) que associa a variável dependente daquele estágio às variáveis independentes de todos os estágios anteriores. A posterior seleção de variáveis independentes apoia-se nos coeficientes da regressão PLS; a cada interação, a variável com menor coeficiente de regressão é removida e um novo modelo PLS é gerado. O erro de predição é então avaliado e uma nova eliminação é promovida até que o número de variáveis remanescentes seja igual ao número de variáveis latentes (condição limite para geração de novos modelos PLS). O conjunto com menor erro determina as variáveis de processo mais relevantes para cada modelo. O conjunto de modelos PLS constituído pelas variáveis selecionadas é então integrado a uma programação quadrática para definição das condições de operação que minimizem o desvio entre os valores preditos e nominais das variáveis de resposta. A sistemática proposta foi validada através de dois exemplos numéricos. O primeiro utilizou dados de uma empresa do setor avícola, enquanto que o segundo apoiou-se em dados simulados. / This dissertation proposes a novel approach for process variable selection and determination of the optimal operating condition in multiple stages, multivariate continuous processes. The proposed framework relies on six steps. First, a pre-treatment of the data is carried out followed by the definition of production stages and removal of outliers. Next, each stage is modeled by a Partial Least Squares regression (PLS) which associates the dependent variable of each stage to all independent variables from previous stages. Independent variables are then iteratively selected based on PLS regression coefficients as follows: the variable with the lowest regression coefficient is removed and a new PLS model is generated. The prediction error is then evaluated and a new elimination is promoted until the number of remaining variables is equal to the number of latent variables (boundary condition for the generation of new PLS models). The subset of independent variables yielding the lowest predictive in each PLS model error is chosen. The set of PLS models consisting of the selected variables is then integrated to a quadratic programming aimed at defining the optimal operating conditions that minimize the deviation between the predicted and nominal values of response variables. The proposed approach was validated through two numerical examples. The first was applied to data from a poultry company, while the second used simulated data.

Esta tese apresenta uma sistemática para seleção de variáveis de processo e determinação da condição ótima de operação em processos contínuos multivariados e em múltiplos estágios. O método proposto é composto por seis etapas. Um pré-tratamento nos dados é realizado após a identificação das variáveis de processo e do estabelecimento dos estágios de produção, onde são descartadas observações com valores espúrios e dados remanescentes são padronizados. Em seguida, cada estágio é modelado através de uma regressão Partial Least Squares (PLS) que associa a variável dependente daquele estágio às variáveis independentes de todos os estágios anteriores. A posterior seleção de variáveis independentes apoia-se nos coeficientes da regressão PLS; a cada interação, a variável com menor coeficiente de regressão é removida e um novo modelo PLS é gerado. O erro de predição é então avaliado e uma nova eliminação é promovida até que o número de variáveis remanescentes seja igual ao número de variáveis latentes (condição limite para geração de novos modelos PLS). O conjunto com menor erro determina as variáveis de processo mais relevantes para cada modelo. O conjunto de modelos PLS constituído pelas variáveis selecionadas é então integrado a uma programação quadrática para definição das condições de operação que minimizem o desvio entre os valores preditos e nominais das variáveis de resposta. A sistemática proposta foi validada através de dois exemplos numéricos. O primeiro utilizou dados de uma empresa do setor avícola, enquanto que o segundo apoiou-se em dados simulados. / This dissertation proposes a novel approach for process variable selection and determination of the optimal operating condition in multiple stages, multivariate continuous processes. The proposed framework relies on six steps. First, a pre-treatment of the data is carried out followed by the definition of production stages and removal of outliers. Next, each stage is modeled by a Partial Least Squares regression (PLS) which associates the dependent variable of each stage to all independent variables from previous stages. Independent variables are then iteratively selected based on PLS regression coefficients as follows: the variable with the lowest regression coefficient is removed and a new PLS model is generated. The prediction error is then evaluated and a new elimination is promoted until the number of remaining variables is equal to the number of latent variables (boundary condition for the generation of new PLS models). The subset of independent variables yielding the lowest predictive in each PLS model error is chosen. The set of PLS models consisting of the selected variables is then integrated to a quadratic programming aimed at defining the optimal operating conditions that minimize the deviation between the predicted and nominal values of response variables. The proposed approach was validated through two numerical examples. The first was applied to data from a poultry company, while the second used simulated data.

Esta tese apresenta uma sistemática para seleção de variáveis de processo e determinação da condição ótima de operação em processos contínuos multivariados e em múltiplos estágios. O método proposto é composto por seis etapas. Um pré-tratamento nos dados é realizado após a identificação das variáveis de processo e do estabelecimento dos estágios de produção, onde são descartadas observações com valores espúrios e dados remanescentes são padronizados. Em seguida, cada estágio é modelado através de uma regressão Partial Least Squares (PLS) que associa a variável dependente daquele estágio às variáveis independentes de todos os estágios anteriores. A posterior seleção de variáveis independentes apoia-se nos coeficientes da regressão PLS; a cada interação, a variável com menor coeficiente de regressão é removida e um novo modelo PLS é gerado. O erro de predição é então avaliado e uma nova eliminação é promovida até que o número de variáveis remanescentes seja igual ao número de variáveis latentes (condição limite para geração de novos modelos PLS). O conjunto com menor erro determina as variáveis de processo mais relevantes para cada modelo. O conjunto de modelos PLS constituído pelas variáveis selecionadas é então integrado a uma programação quadrática para definição das condições de operação que minimizem o desvio entre os valores preditos e nominais das variáveis de resposta. A sistemática proposta foi validada através de dois exemplos numéricos. O primeiro utilizou dados de uma empresa do setor avícola, enquanto que o segundo apoiou-se em dados simulados. / This dissertation proposes a novel approach for process variable selection and determination of the optimal operating condition in multiple stages, multivariate continuous processes. The proposed framework relies on six steps. First, a pre-treatment of the data is carried out followed by the definition of production stages and removal of outliers. Next, each stage is modeled by a Partial Least Squares regression (PLS) which associates the dependent variable of each stage to all independent variables from previous stages. Independent variables are then iteratively selected based on PLS regression coefficients as follows: the variable with the lowest regression coefficient is removed and a new PLS model is generated. The prediction error is then evaluated and a new elimination is promoted until the number of remaining variables is equal to the number of latent variables (boundary condition for the generation of new PLS models). The subset of independent variables yielding the lowest predictive in each PLS model error is chosen. The set of PLS models consisting of the selected variables is then integrated to a quadratic programming aimed at defining the optimal operating conditions that minimize the deviation between the predicted and nominal values of response variables. The proposed approach was validated through two numerical examples. The first was applied to data from a poultry company, while the second used simulated data.

Cette thèse présente de nouveaux modèles et de nouveaux résultats en théorie de la ruine, lorsque les distributions des montants de sinistres sont à queue épaisse. Les hypothèses classiques d’indépendance et de stationnarité, ainsi que l’analyse univariée sont parfois jugées trop restrictives pour décrire l’évolution complexe des réserves d’une compagnie d’assurance. Dans un contexte de dépendance entre les montants de sinistres, des équivalents de la probabilité deruine univariée en temps fini sont obtenus. Cette dépendance, ainsi que les autres paramètres du modèle sont modulés par un processus Markovien d’environnement pour prendre en compte des possibles crises de corrélation. Nous introduisons ensuite des modèles de dépendance entre les montants de sinistres et les temps inter-sinistres pour des risques de type tremblements de terre et inondations. Dans un cadre multivarié, nous présentons divers critères de risques tels que la probabilité de ruine multivariée ou l’espérance de l’intégrale temporelle de la partie négative du processus de risque. Nous résolvons des problèmes d’allocation optimale pour ces différentes mesures de risque. Nous étudions alors l’impact de la dangerosité des risques et de la dépendance entre les branches sur cette allocation optimale / This PhD thesis presents new models and new results in ruin theory, in the case where claim amounts are heavy-tailed distributed. Classical assumptions like independence and stationarity and univariate analysis are sometimes too restrictive to describe the complex evolution of the reserves of an insurance company. In a dependence context, asymptotics of univariate finite-time ruin probability are computed. This dependence, and the other model parameters are modulated by a Markovian environment process to take into account possible correlation crisis. Then, we introduce some models which describe dependence between claim amounts and claim interarrival times we can find in earthquake or flooding risks. In multivariate framework, we present some risk criteria like multivariate ruin probability or the expectation of the timeintegrated negative part of the risk process. We solve some problems of optimal allocation for these risk measures. Then, we study the impact of the risk dangerousness and of the dependence between lines on this optimal allocation.