The main topic of this thesis is how to combine model outputs from deterministic models with measurements from monitoring stations for air pollutants or other meteorological variables. We consider two different approaches to address this particular problem. The first approach is by using the Bayesian Melding (BM) model proposed by Fuentes and Raftery (2005). We successfully implement this model and conduct several simulation studies to examine the performance of this model in different scenarios. We also apply the melding model to the ozone data to show the importance of using the Bayesian melding model to calibrate the model outputs. That is, to adjust the model outputs for the prediction of measurements. Due to the Bayesian framework of the melding model, we can extend it to incorporate other components such as ensemble models, reversible jump MCMC for variable selection. However, the BM model is purely a spatial model and we generally have to deal with space-time dataset in practice. The deficiency of the BM approach leads us to a second approach, an alternative to the BM model, which is a linear mixed model (different from most linear mixed models, the random effects being spatially correlated) with temporally and spatially correlated residuals. We assume the spatial and temporal correlation are separable and use an AR process to model the temporal correlation. We also develop a multivariate version of this model. Both the melding model and linear mixed model are Bayesian hierarchical models, which can better estimate the uncertainties of the estimates and predictions.

The main topic of this thesis is how to combine model outputs from deterministic models with measurements from monitoring stations for air pollutants or other meteorological variables. We consider two different approaches to address this particular problem. The first approach is by using the Bayesian Melding (BM) model proposed by Fuentes and Raftery (2005). We successfully implement this model and conduct several simulation studies to examine the performance of this model in different scenarios. We also apply the melding model to the ozone data to show the importance of using the Bayesian melding model to calibrate the model outputs. That is, to adjust the model outputs for the prediction of measurements. Due to the Bayesian framework of the melding model, we can extend it to incorporate other components such as ensemble models, reversible jump MCMC for variable selection. However, the BM model is purely a spatial model and we generally have to deal with space-time dataset in practice. The deficiency of the BM approach leads us to a second approach, an alternative to the BM model, which is a linear mixed model (different from most linear mixed models, the random effects being spatially correlated) with temporally and spatially correlated residuals. We assume the spatial and temporal correlation are separable and use an AR process to model the temporal correlation. We also develop a multivariate version of this model. Both the melding model and linear mixed model are Bayesian hierarchical models, which can better estimate the uncertainties of the estimates and predictions.

The main topic of this thesis is how to combine model outputs from deterministic models with measurements from monitoring stations for air pollutants or other meteorological variables. We consider two different approaches to address this particular problem. The first approach is by using the Bayesian Melding (BM) model proposed by Fuentes and Raftery (2005). We successfully implement this model and conduct several simulation studies to examine the performance of this model in different scenarios. We also apply the melding model to the ozone data to show the importance of using the Bayesian melding model to calibrate the model outputs. That is, to adjust the model outputs for the prediction of measurements. Due to the Bayesian framework of the melding model, we can extend it to incorporate other components such as ensemble models, reversible jump MCMC for variable selection. However, the BM model is purely a spatial model and we generally have to deal with space-time dataset in practice. The deficiency of the BM approach leads us to a second approach, an alternative to the BM model, which is a linear mixed model (different from most linear mixed models, the random effects being spatially correlated) with temporally and spatially correlated residuals. We assume the spatial and temporal correlation are separable and use an AR process to model the temporal correlation. We also develop a multivariate version of this model. Both the melding model and linear mixed model are Bayesian hierarchical models, which can better estimate the uncertainties of the estimates and predictions. / Science, Faculty of / Statistics, Department of / Graduate

Legacy cyber-physical systems (CPSs) were designed without considering cybersecurity as a primary design tenet especially when considering their evolving operating environment. There are many examples of legacy systems including automotive control, navigation, transportation, and industrial control systems (ICSs), to name a few. To make matters worse, the cost of designing and deploying defenses in existing legacy infrastructure can be overwhelming as millions or even billions of legacy CPS systems are already in use. This economic angle, prevents the use of defenses that are not backward compatible. Moreover, any protection has to operate efficiently in resource constraint environments that are dynamic nature. Hence, the existing approaches that require ex- pensive additional hardware, propose a new protocol from scratch, or rely on complex numerical operations such as strong cryptographic solutions, are less likely to be deployed in practice. In this dissertation, we explore a variety of lightweight solutions for securing different existing CPSs without requiring any modifications to the original system design at hardware or protocol level. In particular, we use fingerprinting, crowdsourcing and deterministic models as alternative backwards- compatible defenses for securing vehicles, global positioning system (GPS) receivers, and a class of ICSs called supervisory control and data acquisition (SCADA) systems, respectively. We use fingerprinting to address the deficiencies in automobile cyber-security from the angle of controller area network (CAN) security. CAN protocol is the de-facto bus standard commonly used in the automotive industry for connecting electronic control units (ECUs) within a vehicle. The broadcast nature of this protocol, along with the lack of authentication or integrity guarantees, create a foothold for adversaries to perform arbitrary data injection or modification and impersonation attacks on the ECUs. We propose SIMPLE, a single-frame based physical layer identification for intrusion detection and prevention on such networks. Physical layer identification or fingerprinting is a method that takes advantage of the manufacturing inconsistencies in the hardware components that generate the analog signal for the CPS of our interest. It translates the manifestation of these inconsistencies, which appear in the analog signals, into unique features called fingerprints which can be used later on for authentication purposes. Our solution is resilient to ambient temperature, supply voltage value variations, or aging. Next, we use fingerprinting and crowdsourcing at two separate protection approaches leveraging two different perspectives for securing GPS receivers against spoofing attacks. GPS, is the most predominant non-authenticated navigation system. The security issues inherent into civilian GPS are exacerbated by the fact that its design and implementation are public knowledge. To address this problem, first we introduce Spotr, a GPS spoofing detection via device fingerprinting, that is able to determine the authenticity of signals based on their physical-layer similarity to the signals that are known to have originated from GPS satellites. More specifically, we are able to detect spoofing activities and track genuine signals over different times and locations and propagation effects related to environmental conditions. In a different approach at a higher level, we put forth Crowdsourcing GPS, a total solution for GPS spoofing detection, recovery and attacker localization. Crowdsourcing is a method where multiple entities share their observations of the environment and get together as a whole to make a more accurate or reliable decision on the status of the system. Crowdsourcing has the advantage of deployment with the less complexity and distributed cost, however its functionality is dependent on the adoption rate by the users. Here, we have two methods for implementing Crowdsourcing GPS. In the first method, the users in the crowd are aware of their approximate distance from other users using Bluetooth. They cross validate this approximate distance with the GPS-derived distance and in case of any discrepancy they report ongoing spoofing activities. This method is a strong candidate when the users in the crowd have a sparse distribution. It is also very effective when tackling multiple coordinated adversaries. For method II, we exploit the angular dispersion of the users with respect to the direction that the adversarial signal is being transmitted from. As a result, the users that are not facing the attacker will be safe. The reason for this is that human body mostly comprises of water and absorbs the weak adversarial GPS signal. The safe users will help the spoofed users find out that there is an ongoing attack and recover from it. Additionally, the angular information is used for localizing the adversary. This method is slightly more complex, and shows the best performance in dense areas. It is also designed based on the assumption that the spoofing attack is only terrestrial. Finally, we propose a tandem IDS to secure SCADA systems. SCADA systems play a critical role in most safety-critical infrastructures of ICSs. The evolution of communications technology has rendered modern SCADA systems and their connecting actuators and sensors vulnerable to malicious attacks on both physical and application layers. The conventional IDS that are built for securing SCADA systems are focused on a single layer of the system. With the tandem IDS we break this habit and propose a strong multi-layer solution which is able to expose a wide range of attack. To be more specific, the tandem IDS comprises of two parts, a traditional network IDS and a shadow replica. We design the shadow replica as a deterministic IDS. It performs a workflow analysis and makes sure the logical flow of the events in the SCADA controller and its connected devices maintain their expected states. Any deviation would be a malicious activity or a reliability issue. To model the application level events, we leverage finite state machines (FSMs) to compute the anticipated states of all of the devices. This is feasible because in many of the existing ICSs the flow of traffic and the resulting states and actions in the connected devices have a deterministic nature. Consequently, it leads to a reliable and free of uncertainty solution. Aside from detecting traditional network attacks, our approach bypasses the attacker in case it succeeds in taking over the devices and also maintains continuous service if the SCADA controller gets compromised. / Doctor of Philosophy / Our lives are entangled with cyber-physical systems (CPSs) on a daily basis. Examples of these systems are vehicles, navigation systems, transportation systems, industrial control systems, etc. CPSs are mostly legacy systems and were built with a focus on performance, overlooking security. Security was not considered in the design of these old systems and now they are dominantly used in our everyday life. After numerous demonstration of cyber hacks, the necessity of protecting the CPSs from adversarial activities is no longer ambiguous. Many of the advanced cryptographic techniques are far too complex to be implemented in the existing CPSs such as cars, satellites, etc. We attempt to secure such resource constraint systems using simple backward compatible techniques in this dissertation. We design cheap lightweight solutions, with no modifications to the original system. In part of our research, we use fingerprinting as a technique to secure passenger cars from being hacked, and GPS receivers from being spoofed. For a brief description of fingerprinting, we use the example of two identical T-shirts with the same size and design. They will always have subtle differences between them no matter how hard the tailor tried to make them identical. This means that there are no two T-shirts that are exactly identical. This idea, when applied to analog signalling on electric devices, is called fingerprinting. Here, we fingerprint the mini computers inside a car, which enables us to identify these computers and prevent hacking. We also use the signal levels to design fingerprints for GPS signals. We use the fingerprints to distinguish counterfeit GPS signals from the ones that have originated from genuine satellites. This summarizes two major contributions in the dissertation. Our earlier contribution to GPS security was effective, but it was heavily dependent on the underlying hardware, requiring extensive training for each radio receiver that it was protecting. To remove this dependence of training for the specific underlying hardware, we design and implement the next framework using defenses that require application-layer access. Thus, we proposed two methods that leverage crowdsourcing approaches to defend against GPS spoofing attacks and, at the same time, improve the accuracy of localization for commodity mobile devices. Crowdsourcing is a method were several devices agree to share their information with each other. In this work, GPS users share their location and direction information, and in case of any discrepancy they figure that they are under attack and cooperate to recover from it. Last, we shift the gear to the industrial control systems (ICSs) and propose a novel IDS to protect them against various cyber attacks. Unlike the conventional IDSs that are focused on one of the layers of the system, our IDS comprises of two main components. A conventional component that exposes traditional attacks and a second component called a shadow replica. The replica mimics the behavior of the system and compares it with that of the actual system in a real-time manner. In case of any deviation between the two, it detects attacks that target the logical flow of the events in the system. Note that such attacks are more sophisticated and difficult to detect because they do not leave any obvious footprints behind. Upon detection of attacks on the original controller, our replica takes over the responsibilities of the original ICS controller and provides service continuity.

Contexto: As infecções causadas pelo Staphylococcus aureus resistente a meticilina tem conseqüências graves aos pacientes e aos serviços de saúde. As medidas gerais de controle de infecção são fundamentais para conter a disseminação desta bactéria, mas não são sempre suficientes, e estratégias mais agressivas são empregadas muitas vezes para obter-se uma diminuição das taxas. Os estudos publicados para a avaliação da efetividade das medidas tem recebido criticas pela sua qualidade, e ainda não ha consenso sobre o assunto. Objetivos: Quantificar a prevalência de pacientes colonizados pelo Staphylococcus aureus resistentes a meticilina admitidos em um hospital terciário universitário, avaliar aspectos associados a transmissão da bactéria e estimar, através de modelo matemático, o efeito de medidas de isolamento de pacientes com Staphylococcus aureus resistentes meticilina na aquisic5o de colonização. Métodos: Um estudo de coorte prospectivo, com pacientes clínicos admitidos no hospital selecionados aleatoriamente na sua admissão. Os pacientes foram entrevistados na admissão e a cada semana e houve coleta de swab nasal para identificação de colonização pelo MRSA na admissão e a cada semana de internação. Os fatores de risco associados com a presença de colonização na admissão e sua aquisição durante a hospitalização foram investigados. Construiu-se um modelo determinístico de aquisição de MRSA considerando-se o impacto da prevalência de colonização basal pelo MRSA, a adesão a higienização das mãos e o use de isolamento na carga de trabalhos de enfermeiros e nas taxas de infecção pelo MRSA. Resultados: Foram selecionados 301 pacientes adultos e 189 pediátricos. A prevalência de MRSA na admissão foi de 5,3% (IC 95%: 3,1% a 8,5%) na população adulta, e 1,6% (1C95%: 0,3% a 4,6%) entre as crianças. A analise multivariável identificou os seguintes fatores de risco associados com a colonização em adultos na admissão: idade acima de 60 anos (RP = 2,9, IC 95%: 1,1 a 7,6), e internação no ano anterior (RP = 5,4, IC 95%: 1,3 a 23,4). A analise da incidência foi realizada em 285 pacientes que não eram portadores de MRSA na admissão, e que tiveram uma segunda amostra coletada. Entre estes, 9,5% (1C95%: 8,1 a 11,1) adquiriram colonização ou infecção durante a hospitalização. A taxa de colonização foi de 5,5/1.000 pacientes-dia entre os adultos (1C95%: 3,36 a 8,49), e de 1,6/1.000 pacientes-dia entre as crianças (1C95%: 0,33 a 4,61). 0 risco de aquisição aumentou conforme aumentou o tempo de internação, e este aumento foi estatisticamente significativo. No modelo proposto, se isolamento de pacientes aumenta o volume de trabalho dos trabalhadores da saúde, reduzindo a adesão aos protocolos básicos de prevenção de infecção, a execução das medidas de isolamento podem provocar o aumento da prevalência de MRSA. Entretanto, a aplicação de medidas de isolamento pode ter resultados diferentes dependendo da prevalência inicial MRSA e da carga de trabalho dos trabalhadores. A redução da carga de trabalho é mais custo-efetivo quando implementada para impedir a transição de um estado de baixa endemicidade para de alta. Conclusões: A identificação de portadores nasais de MRSA poderia ajudar a diminuir a pressão de colonização dentro do hospital e reduzir o alto índice de aquisição nesses pacientes. Entretanto, as medidas de controle devem ser cuidadosamente avaliadas num cenário com altas taxas de infecção, pois o grande numero de pacientes em isolamento pode comprometer a adesão a higienização de mãos ao aumentar a carga de trabalho dos enfermeiros, tendo efeito contrario ao desejado na prevenção de infecções. / Background: Infections caused by methicillin-resistant Staphylococcus aureus have serious consequences for patients and healthcare services. Basic infection control measures are essential to contain the spread of the bacteria, but are not always sufficient, and more aggressive strategies are often employed to obtain a reduction of rates. Studies published to assess the effectiveness of the measures have received criticism for its quality, and there is no consensus if the search and destroy policy is necessary Objectives: To quantify the prevalence of patients colonized by methicillin-resistant Staphylococcus aureus admitted in a tertiary university hospital, evaluate factors associated with transmission of the bacteria and to estimate, through a deterministic mathematical model, the effect of isolation measures of patients with the bacteria in the acquisition of colonization. Methods: A prospective cohort with clinical patients randomly selected in their entry to a hospital. Patients were interviewed on admission and weekly, with collection of nasal swab for identification of MRSA colonization at admission and every week of hospitalization. Risk factors associated with the presence of colonization on admission, and associated to its acquisition during hospitalization were investigated. A deterministic model for acquisition of MRSA was built, considering the impact of the baseline prevalence of MRSA colonization, adherence to hand hygiene and use of isolation measures in the workload healthcare workers and on the rates of MRSA infection. Results: We randomly selected 301 adult and 189 pediatric patients. The prevalence of MRSA at admission was 5.3% (95% CI, 3.1% to 8.5%) in adults and 1.6% (95% CI, 0.3% to 4.6%) among children. Multivariate analysis identified the following risk factors associated with colonization of adults at admission: age over 60 years (PR = 2.9 95% CI 1.1 10 7.6) and hospitalization in the previous year (PR = 5.4, 95% CI 1.3 to 23.4). The analysis of incidence was performed in 285 patients who were not carriers of MRSA on admission, and who had a second sample collected. Of these, 9.5% (95% CI 8.1 to 11.1) acquired colonization or infection during hospitalization. The colonization rate was 5.5 / 1,000 patient-days among adults (1C95%, from 3.36 to 8.49) and 1.6 / 1,000 patient-days among children (95% Cl, 0.33 to 4.61). The risk of acquisition increases as the length of stay increases, and this increase is statistically significant. In the proposed model, if isolation of patients increases the workload of health workers, and reduces adherence to the basic protocols for infection prevention, the implementation isolation can increase the prevalence of MRSA. However, the application of isolation can have different results depending on the initial prevalence of MRSA and the effect on healthcare workers workload. The reduction in workload is more cost-effective when implemented to prevent the transition from a state of low to hitzh. endemicity. Conclusions: The identification of nasal carriers of MRSA could help reduce the colonization pressure within the hospital and reduce the high rate of acquisition in these patients, but the measures of control should be carefully evaluated in a setting with high rates of infection, because the great number of patients in isolation can compromise adherence to hand hygiene of increasing the workload of the nurses, and result in the opposite effect in preventing infections.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Nesta dissertação, propõe-se uma metodologia para a analise da resposta dinâmica de pontes rodoviárias devido à travessia de comboios de diversos tipos de veículos sobre o tabuleiro irregular dessas obras de arte. Para tal, desenvolve-se uma análise paramétrica extensa com o objetivo de avaliarem-se os efeitos dinâmicos provenientes das irregularidades superficiais existentes no tabuleiro sobre o comportamento das pontes rodoviárias. A metodologia de análise é desenvolvida no domínio do tempo de acordo com um modelo estatístico. O modelo matemático é concebido de forma a simular o conjunto dos veículos e do tabuleiro, denominado neste trabalho comumente de sistema veículo-ponte (ou sistema veículo-estrutura). Considera-se a participação da massa e da rigidez dos veículos na definição das freqüências do conjunto e, conseqüentemente, a força de interação entre os veículos e a ponte é afetada pela flexibilidade desta. Simula-se o tabuleiro das obras de arte por uma viga modelada com base em elementos finitos de barra unidimensionais, com massas concentradas em seus nós e flexibilidade distribuída. Aos nós estão associados os movimentos de rotação no plano e de translação vertical. Desprezam-se a inércia de rotação e a deformação por cisalhamento. São considerados 04 (quatro) modelos distintos para representar os veículos do comboio na análise paramétrica, sendo estes: veículos com um eixo e uma massa, viaturas com um eixo e duas massas, veículos com dois eixos e três massas e carros com três eixos e quatro massas. Todos os veículos são simulados por sistemas de massas, molas e amortecedores e são descritos por graus de liberdade à translação e rotação no plano. As irregularidades da pista são definidas por um modelo matemático não- determinístico, com base na densidade espectral do perfil do pavimento, obtida experimentalmente. Os perfis irregulares do pavimento são considerados associados a processos fracamente estacionários e ergódicos. O carregamento sobre a superestrutura das pontes é constituído por uma sucessão infinita de veículos, igualmente espaçados e deslocando-se com velocidade constante sobre o tabuleiro. Devido à própria natureza das irregularidades superficiais e do comboio de veículos, atenção especial é concentrada na fase permanente da resposta do sistema veículo-ponte. São estudadas as respostas dos modelos estruturais, com base em tabuleiros isostáticos de concreto armado, com e sem balanços, em seção do tipo caixão, em termos de deslocamentos e esforços nas seções onde ocorrem os efeitos máximos. As conclusões deste trabalho versam sobre a adequação do modelo matemático empregado, observando- se a influência do tipo de modelo de veículo utilizado (com um ou mais eixos) sobre aresposta dinâmica das pontes rodoviárias analisadas e, bem como, à magnitude dos efeitos dinâmicos provenientes das irregularidades superficiais e suas conseqüências sobre as atitudes correntes de projeto. / In this investigation an analysis methodology is developed to evaluate the vehicle- structure response. A parametric study is carried to evaluate the dynamical effects, displacements and stresses, on highway bridge decks, due to vehicles crossing on rough pavement surfaces defined by a probabilistic model. The analysis methodology was considered following a statistical model, in the time domain. The mathematical model assumes a finite element representation of a beam like deck and the vehicle simulation uses concentrated parameters of mass, stiffness and damping. Four different models are developed in order to represent the vehicles. The vehicles are modeled as one axle, with one or two masses, and two and three axles, with three and four masses, respectively. All vehicles are simulated as mass-spring-damper systems and the degrees of freedom of these cars are defined as in plane vertical translations and rotations. The deck surface roughness is defined by a well known power spectrum probability density of road pavement profiles. The irregular pavement surface was defined like a weakly stationary and ergodic random process. The moving load is formed by an infinite succession of vehicles moving with constant velocity and equally spaced. Only steady-state response is considered. Response data are produced on concrete box girder elements assembled as simple beams, including cantilever spans. Conclusions are concerned with the fitness of the developed analysis methodology and the magnitude of the response amplification due to the surface irregularities.

Contexto: As infecções causadas pelo Staphylococcus aureus resistente a meticilina tem conseqüências graves aos pacientes e aos serviços de saúde. As medidas gerais de controle de infecção são fundamentais para conter a disseminação desta bactéria, mas não são sempre suficientes, e estratégias mais agressivas são empregadas muitas vezes para obter-se uma diminuição das taxas. Os estudos publicados para a avaliação da efetividade das medidas tem recebido criticas pela sua qualidade, e ainda não ha consenso sobre o assunto. Objetivos: Quantificar a prevalência de pacientes colonizados pelo Staphylococcus aureus resistentes a meticilina admitidos em um hospital terciário universitário, avaliar aspectos associados a transmissão da bactéria e estimar, através de modelo matemático, o efeito de medidas de isolamento de pacientes com Staphylococcus aureus resistentes meticilina na aquisic5o de colonização. Métodos: Um estudo de coorte prospectivo, com pacientes clínicos admitidos no hospital selecionados aleatoriamente na sua admissão. Os pacientes foram entrevistados na admissão e a cada semana e houve coleta de swab nasal para identificação de colonização pelo MRSA na admissão e a cada semana de internação. Os fatores de risco associados com a presença de colonização na admissão e sua aquisição durante a hospitalização foram investigados. Construiu-se um modelo determinístico de aquisição de MRSA considerando-se o impacto da prevalência de colonização basal pelo MRSA, a adesão a higienização das mãos e o use de isolamento na carga de trabalhos de enfermeiros e nas taxas de infecção pelo MRSA. Resultados: Foram selecionados 301 pacientes adultos e 189 pediátricos. A prevalência de MRSA na admissão foi de 5,3% (IC 95%: 3,1% a 8,5%) na população adulta, e 1,6% (1C95%: 0,3% a 4,6%) entre as crianças. A analise multivariável identificou os seguintes fatores de risco associados com a colonização em adultos na admissão: idade acima de 60 anos (RP = 2,9, IC 95%: 1,1 a 7,6), e internação no ano anterior (RP = 5,4, IC 95%: 1,3 a 23,4). A analise da incidência foi realizada em 285 pacientes que não eram portadores de MRSA na admissão, e que tiveram uma segunda amostra coletada. Entre estes, 9,5% (1C95%: 8,1 a 11,1) adquiriram colonização ou infecção durante a hospitalização. A taxa de colonização foi de 5,5/1.000 pacientes-dia entre os adultos (1C95%: 3,36 a 8,49), e de 1,6/1.000 pacientes-dia entre as crianças (1C95%: 0,33 a 4,61). 0 risco de aquisição aumentou conforme aumentou o tempo de internação, e este aumento foi estatisticamente significativo. No modelo proposto, se isolamento de pacientes aumenta o volume de trabalho dos trabalhadores da saúde, reduzindo a adesão aos protocolos básicos de prevenção de infecção, a execução das medidas de isolamento podem provocar o aumento da prevalência de MRSA. Entretanto, a aplicação de medidas de isolamento pode ter resultados diferentes dependendo da prevalência inicial MRSA e da carga de trabalho dos trabalhadores. A redução da carga de trabalho é mais custo-efetivo quando implementada para impedir a transição de um estado de baixa endemicidade para de alta. Conclusões: A identificação de portadores nasais de MRSA poderia ajudar a diminuir a pressão de colonização dentro do hospital e reduzir o alto índice de aquisição nesses pacientes. Entretanto, as medidas de controle devem ser cuidadosamente avaliadas num cenário com altas taxas de infecção, pois o grande numero de pacientes em isolamento pode comprometer a adesão a higienização de mãos ao aumentar a carga de trabalho dos enfermeiros, tendo efeito contrario ao desejado na prevenção de infecções. / Background: Infections caused by methicillin-resistant Staphylococcus aureus have serious consequences for patients and healthcare services. Basic infection control measures are essential to contain the spread of the bacteria, but are not always sufficient, and more aggressive strategies are often employed to obtain a reduction of rates. Studies published to assess the effectiveness of the measures have received criticism for its quality, and there is no consensus if the search and destroy policy is necessary Objectives: To quantify the prevalence of patients colonized by methicillin-resistant Staphylococcus aureus admitted in a tertiary university hospital, evaluate factors associated with transmission of the bacteria and to estimate, through a deterministic mathematical model, the effect of isolation measures of patients with the bacteria in the acquisition of colonization. Methods: A prospective cohort with clinical patients randomly selected in their entry to a hospital. Patients were interviewed on admission and weekly, with collection of nasal swab for identification of MRSA colonization at admission and every week of hospitalization. Risk factors associated with the presence of colonization on admission, and associated to its acquisition during hospitalization were investigated. A deterministic model for acquisition of MRSA was built, considering the impact of the baseline prevalence of MRSA colonization, adherence to hand hygiene and use of isolation measures in the workload healthcare workers and on the rates of MRSA infection. Results: We randomly selected 301 adult and 189 pediatric patients. The prevalence of MRSA at admission was 5.3% (95% CI, 3.1% to 8.5%) in adults and 1.6% (95% CI, 0.3% to 4.6%) among children. Multivariate analysis identified the following risk factors associated with colonization of adults at admission: age over 60 years (PR = 2.9 95% CI 1.1 10 7.6) and hospitalization in the previous year (PR = 5.4, 95% CI 1.3 to 23.4). The analysis of incidence was performed in 285 patients who were not carriers of MRSA on admission, and who had a second sample collected. Of these, 9.5% (95% CI 8.1 to 11.1) acquired colonization or infection during hospitalization. The colonization rate was 5.5 / 1,000 patient-days among adults (1C95%, from 3.36 to 8.49) and 1.6 / 1,000 patient-days among children (95% Cl, 0.33 to 4.61). The risk of acquisition increases as the length of stay increases, and this increase is statistically significant. In the proposed model, if isolation of patients increases the workload of health workers, and reduces adherence to the basic protocols for infection prevention, the implementation isolation can increase the prevalence of MRSA. However, the application of isolation can have different results depending on the initial prevalence of MRSA and the effect on healthcare workers workload. The reduction in workload is more cost-effective when implemented to prevent the transition from a state of low to hitzh. endemicity. Conclusions: The identification of nasal carriers of MRSA could help reduce the colonization pressure within the hospital and reduce the high rate of acquisition in these patients, but the measures of control should be carefully evaluated in a setting with high rates of infection, because the great number of patients in isolation can compromise adherence to hand hygiene of increasing the workload of the nurses, and result in the opposite effect in preventing infections.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Nesta dissertação, propõe-se uma metodologia para a analise da resposta dinâmica de pontes rodoviárias devido à travessia de comboios de diversos tipos de veículos sobre o tabuleiro irregular dessas obras de arte. Para tal, desenvolve-se uma análise paramétrica extensa com o objetivo de avaliarem-se os efeitos dinâmicos provenientes das irregularidades superficiais existentes no tabuleiro sobre o comportamento das pontes rodoviárias. A metodologia de análise é desenvolvida no domínio do tempo de acordo com um modelo estatístico. O modelo matemático é concebido de forma a simular o conjunto dos veículos e do tabuleiro, denominado neste trabalho comumente de sistema veículo-ponte (ou sistema veículo-estrutura). Considera-se a participação da massa e da rigidez dos veículos na definição das freqüências do conjunto e, conseqüentemente, a força de interação entre os veículos e a ponte é afetada pela flexibilidade desta. Simula-se o tabuleiro das obras de arte por uma viga modelada com base em elementos finitos de barra unidimensionais, com massas concentradas em seus nós e flexibilidade distribuída. Aos nós estão associados os movimentos de rotação no plano e de translação vertical. Desprezam-se a inércia de rotação e a deformação por cisalhamento. São considerados 04 (quatro) modelos distintos para representar os veículos do comboio na análise paramétrica, sendo estes: veículos com um eixo e uma massa, viaturas com um eixo e duas massas, veículos com dois eixos e três massas e carros com três eixos e quatro massas. Todos os veículos são simulados por sistemas de massas, molas e amortecedores e são descritos por graus de liberdade à translação e rotação no plano. As irregularidades da pista são definidas por um modelo matemático não- determinístico, com base na densidade espectral do perfil do pavimento, obtida experimentalmente. Os perfis irregulares do pavimento são considerados associados a processos fracamente estacionários e ergódicos. O carregamento sobre a superestrutura das pontes é constituído por uma sucessão infinita de veículos, igualmente espaçados e deslocando-se com velocidade constante sobre o tabuleiro. Devido à própria natureza das irregularidades superficiais e do comboio de veículos, atenção especial é concentrada na fase permanente da resposta do sistema veículo-ponte. São estudadas as respostas dos modelos estruturais, com base em tabuleiros isostáticos de concreto armado, com e sem balanços, em seção do tipo caixão, em termos de deslocamentos e esforços nas seções onde ocorrem os efeitos máximos. As conclusões deste trabalho versam sobre a adequação do modelo matemático empregado, observando- se a influência do tipo de modelo de veículo utilizado (com um ou mais eixos) sobre aresposta dinâmica das pontes rodoviárias analisadas e, bem como, à magnitude dos efeitos dinâmicos provenientes das irregularidades superficiais e suas conseqüências sobre as atitudes correntes de projeto. / In this investigation an analysis methodology is developed to evaluate the vehicle- structure response. A parametric study is carried to evaluate the dynamical effects, displacements and stresses, on highway bridge decks, due to vehicles crossing on rough pavement surfaces defined by a probabilistic model. The analysis methodology was considered following a statistical model, in the time domain. The mathematical model assumes a finite element representation of a beam like deck and the vehicle simulation uses concentrated parameters of mass, stiffness and damping. Four different models are developed in order to represent the vehicles. The vehicles are modeled as one axle, with one or two masses, and two and three axles, with three and four masses, respectively. All vehicles are simulated as mass-spring-damper systems and the degrees of freedom of these cars are defined as in plane vertical translations and rotations. The deck surface roughness is defined by a well known power spectrum probability density of road pavement profiles. The irregular pavement surface was defined like a weakly stationary and ergodic random process. The moving load is formed by an infinite succession of vehicles moving with constant velocity and equally spaced. Only steady-state response is considered. Response data are produced on concrete box girder elements assembled as simple beams, including cantilever spans. Conclusions are concerned with the fitness of the developed analysis methodology and the magnitude of the response amplification due to the surface irregularities.

Contexto: As infecções causadas pelo Staphylococcus aureus resistente a meticilina tem conseqüências graves aos pacientes e aos serviços de saúde. As medidas gerais de controle de infecção são fundamentais para conter a disseminação desta bactéria, mas não são sempre suficientes, e estratégias mais agressivas são empregadas muitas vezes para obter-se uma diminuição das taxas. Os estudos publicados para a avaliação da efetividade das medidas tem recebido criticas pela sua qualidade, e ainda não ha consenso sobre o assunto. Objetivos: Quantificar a prevalência de pacientes colonizados pelo Staphylococcus aureus resistentes a meticilina admitidos em um hospital terciário universitário, avaliar aspectos associados a transmissão da bactéria e estimar, através de modelo matemático, o efeito de medidas de isolamento de pacientes com Staphylococcus aureus resistentes meticilina na aquisic5o de colonização. Métodos: Um estudo de coorte prospectivo, com pacientes clínicos admitidos no hospital selecionados aleatoriamente na sua admissão. Os pacientes foram entrevistados na admissão e a cada semana e houve coleta de swab nasal para identificação de colonização pelo MRSA na admissão e a cada semana de internação. Os fatores de risco associados com a presença de colonização na admissão e sua aquisição durante a hospitalização foram investigados. Construiu-se um modelo determinístico de aquisição de MRSA considerando-se o impacto da prevalência de colonização basal pelo MRSA, a adesão a higienização das mãos e o use de isolamento na carga de trabalhos de enfermeiros e nas taxas de infecção pelo MRSA. Resultados: Foram selecionados 301 pacientes adultos e 189 pediátricos. A prevalência de MRSA na admissão foi de 5,3% (IC 95%: 3,1% a 8,5%) na população adulta, e 1,6% (1C95%: 0,3% a 4,6%) entre as crianças. A analise multivariável identificou os seguintes fatores de risco associados com a colonização em adultos na admissão: idade acima de 60 anos (RP = 2,9, IC 95%: 1,1 a 7,6), e internação no ano anterior (RP = 5,4, IC 95%: 1,3 a 23,4). A analise da incidência foi realizada em 285 pacientes que não eram portadores de MRSA na admissão, e que tiveram uma segunda amostra coletada. Entre estes, 9,5% (1C95%: 8,1 a 11,1) adquiriram colonização ou infecção durante a hospitalização. A taxa de colonização foi de 5,5/1.000 pacientes-dia entre os adultos (1C95%: 3,36 a 8,49), e de 1,6/1.000 pacientes-dia entre as crianças (1C95%: 0,33 a 4,61). 0 risco de aquisição aumentou conforme aumentou o tempo de internação, e este aumento foi estatisticamente significativo. No modelo proposto, se isolamento de pacientes aumenta o volume de trabalho dos trabalhadores da saúde, reduzindo a adesão aos protocolos básicos de prevenção de infecção, a execução das medidas de isolamento podem provocar o aumento da prevalência de MRSA. Entretanto, a aplicação de medidas de isolamento pode ter resultados diferentes dependendo da prevalência inicial MRSA e da carga de trabalho dos trabalhadores. A redução da carga de trabalho é mais custo-efetivo quando implementada para impedir a transição de um estado de baixa endemicidade para de alta. Conclusões: A identificação de portadores nasais de MRSA poderia ajudar a diminuir a pressão de colonização dentro do hospital e reduzir o alto índice de aquisição nesses pacientes. Entretanto, as medidas de controle devem ser cuidadosamente avaliadas num cenário com altas taxas de infecção, pois o grande numero de pacientes em isolamento pode comprometer a adesão a higienização de mãos ao aumentar a carga de trabalho dos enfermeiros, tendo efeito contrario ao desejado na prevenção de infecções. / Background: Infections caused by methicillin-resistant Staphylococcus aureus have serious consequences for patients and healthcare services. Basic infection control measures are essential to contain the spread of the bacteria, but are not always sufficient, and more aggressive strategies are often employed to obtain a reduction of rates. Studies published to assess the effectiveness of the measures have received criticism for its quality, and there is no consensus if the search and destroy policy is necessary Objectives: To quantify the prevalence of patients colonized by methicillin-resistant Staphylococcus aureus admitted in a tertiary university hospital, evaluate factors associated with transmission of the bacteria and to estimate, through a deterministic mathematical model, the effect of isolation measures of patients with the bacteria in the acquisition of colonization. Methods: A prospective cohort with clinical patients randomly selected in their entry to a hospital. Patients were interviewed on admission and weekly, with collection of nasal swab for identification of MRSA colonization at admission and every week of hospitalization. Risk factors associated with the presence of colonization on admission, and associated to its acquisition during hospitalization were investigated. A deterministic model for acquisition of MRSA was built, considering the impact of the baseline prevalence of MRSA colonization, adherence to hand hygiene and use of isolation measures in the workload healthcare workers and on the rates of MRSA infection. Results: We randomly selected 301 adult and 189 pediatric patients. The prevalence of MRSA at admission was 5.3% (95% CI, 3.1% to 8.5%) in adults and 1.6% (95% CI, 0.3% to 4.6%) among children. Multivariate analysis identified the following risk factors associated with colonization of adults at admission: age over 60 years (PR = 2.9 95% CI 1.1 10 7.6) and hospitalization in the previous year (PR = 5.4, 95% CI 1.3 to 23.4). The analysis of incidence was performed in 285 patients who were not carriers of MRSA on admission, and who had a second sample collected. Of these, 9.5% (95% CI 8.1 to 11.1) acquired colonization or infection during hospitalization. The colonization rate was 5.5 / 1,000 patient-days among adults (1C95%, from 3.36 to 8.49) and 1.6 / 1,000 patient-days among children (95% Cl, 0.33 to 4.61). The risk of acquisition increases as the length of stay increases, and this increase is statistically significant. In the proposed model, if isolation of patients increases the workload of health workers, and reduces adherence to the basic protocols for infection prevention, the implementation isolation can increase the prevalence of MRSA. However, the application of isolation can have different results depending on the initial prevalence of MRSA and the effect on healthcare workers workload. The reduction in workload is more cost-effective when implemented to prevent the transition from a state of low to hitzh. endemicity. Conclusions: The identification of nasal carriers of MRSA could help reduce the colonization pressure within the hospital and reduce the high rate of acquisition in these patients, but the measures of control should be carefully evaluated in a setting with high rates of infection, because the great number of patients in isolation can compromise adherence to hand hygiene of increasing the workload of the nurses, and result in the opposite effect in preventing infections.

En basant l'analyse des modèles épidémiologiques sur leur représentation graphique plutôt que sur leurs équations différentielles, il est possible de mettre en évidence plusieurs concepts importants à l'aide des composantes d'un hypergraphe. On décrit une manière formelle de créer automatiquement un système d'équations différentielles à partir de ces composantes et on adapte ensuite la définition du produit cartésien pour les hypergraphes décrits, ce qui permet la fusion de modèles. À l'aide d'un algorithme qui ajoute automatiquement de nouvelles composantes à l'hypergraphe, il est possible d'isoler virtuellement certains individus, afin d'expliciter le calcul de nombres de reproduction. On montre ensuite que la forme des équations différentielles créées admettent une solution unique et que l'algorithme d'ajout aux hypergraphes est stable au niveau de la structure et de la dynamique des hypergraphes. On trouve que la méthode décrite pour le calcul des nombres de reproduction permet une meilleure prédiction de la croissance de l'épidémie que le calcul standard \(\mathcal{R}_t = \mathcal{R}_0 \cdot S / N\) et que le calcul de \(\mathcal{R}_0\) est très similaire aux résultats trouvés à l'aide de la matrice de prochaine génération, en plus d'être plus simple à mettre en place et d'offrir une justification plus robuste. On conclue ce mémoire en décrivant sommairement un processus d'apprentissage automatique des paramètres dans les modèles à compartiments, afin de permettre une calibration de modèles plus rapide. L'apprentissage machine peut être intégré en faisant appel à la librarie torchdiffeq, qui implémente les équations différentielles ordinaires neuronales en utilisant Pytorch. / By basing the analysis of epidemiological models on their graphical representation rather than on their differential equations, it is possible to highlight a few key concepts by using the components of a hypergraph. We give a formal way to automatically create a system of differential equations by using these components and we then adapt the definition of the cartesian product for the defined hypergraphs, which permits the merging of models. Using an algorithm which automatically adds new components to the graph, we can virtually isolate a few individuals to explicitly compute the reproduction numbers. We then show that the resulting differential equations allow for a unique solution and that the modification algorithm is stable for the structure and dynamics of the hypergraphs. We find that the described method for the computation of reproduction numbers gives a more accurate prediction of the growth of the epidemic than the standard computation \(\mathcal{R}_t = \mathcal{R}_0 \cdot S/N\) and that the computation of \(\mathcal{R}_0\) is very similar to the results found using the next generation matrix method, as well as being simpler to integrate into models and offering a more robust justification. We conclude this thesis with a brief outline of an automatic learning process for the parameters in compartmental models, which allows a faster calibration of epidemiological models. The implementation of machine learning can be done through the torchdiffeq library, which applies the theory of neural ordinary differential equations using Pytorch.