Information technology (IT) is a key enabler of modern business practices, yet reliably effective IT systems remain a significant challenge for many organizations. The consequences when systems fail to behave as expected becomes ever-more problematic as IT dependence grows. Therefore, methods for assessing IT effectiveness and generating actionable recommendations for improvement are key drivers of success. For this reason, large organizations often adopt IT best practice frameworks such as COBIT, ITIL or ISO/IEC standards which can offer greater assurances of IT effectiveness. However smaller organizations are rarely able to adopt these frameworks due, in part, to resource constraints, and a preference to eschew authoritative practices in favour of informal guides to action. Consequently, a significant research gap is the lack of IT effectiveness approaches for organizations unable or unwilling to adopt formal IT best practice frameworks. This thesis presents an alternative norms-based approach to IT effectiveness which some organizations might find more suitable. Norms are informal beliefs (e.g. ‘using a complex password helps safeguard data’) which motivate behaviours and can often be expressed using non-technical language. We review the literature to formulate a predictive model connecting norms to IT quality. Employing a scientific methodology defensible on philosophical grounds and accepted research practices, we distil a set of IT effectiveness norms from the COBIT 4.1 IT governance framework and adapt theories of motivation to justify our assertion that IT effectiveness norms can motivate actions. Our work is signficant in its formulation of an alternative approach for assessing IT operations and improving organizational IT outcomes. Our survey instrument –validated in four studies, which include a non-profit and government organization, multiple small businesses, a large pharmaceutical company and a university –is a light-weight and reliable assessment tool. Our predictive model is able to explain 26% of observed variance, and can offer actionable and non-technical insights which can improve organizational outcomes. A norms-based approach may bring many of the same IT effectiveness benefits offered by formal IT best practices into organizations, such as small businesses, which lack the resources for their implementation. This approach may also help bridge important communication gaps between IT professionals and others in the organization by providing a different, less technical perspective for framing, assessing, diagnosing, and communicating about IT processes.

Sarcopenia (Sc) é uma doença caracterizada por sintomas patológicos sem causas específicas que acomete parte dos idosos. A Sc promove reduções na massa muscular (MM) e força muscular (FM), com sérios impactos funcionais e motores. A força de preensão manual (FPM) utilizada para diagnosticar Sc não é representativa da FM global, especialmente para FM de membros inferiores (MMII), quando idosos passam por intervenção com treinamento de FM. Embora utilizada para definir a Sc, FM é considerada de forma absoluta ou relativizada pela massa corporal (MC), cuja relação nem sempre é linear. Assim, o objetivo deste estudo foi estabelecer parâmetros válidos para diagnóstico da Sc, a partir de diferentes expressões da FM de MMII em idosos. Uma amostra de 98 idosos fisicamente independentes foi medida pela absorciometria radiológia de dupla energia (DXA) para determinação do Tecido Mole Magro apendicular. Os idosos foram classificados por sexo e Sc (1=Sc; 0=não Sc), segundo os critérios do European Working Group on Sarcopenia in Older People (EWGSOP). A FM de extensão de joelhos (FMEJ) isocinética determinada a 60º/s (Biodex, System 4 Pro), foi considerada como FM referencial (FMEJTorquePico-60°/s), e a FMEJ Estimada em cadeira extensora (CMáxEstFMEJcad-ext), em protocolo de repetição máxima. Análise descritiva com medidas de tendência central foi utilizada para caracterização da amostra. A validação da FMEJ Estimada foi testada por correlação com a FMEJ de referência. Os valores de CMáxEstFMEJcad-ext foram relativizados pela massa corporal (FMEJ/MC) e por ajustes alométricos (FMEJ/MCb), onde b é o expoente gerado por regressão log-linear entre FMEJ e MC. Para determinar um modelo explicativo da Sc a partir de cada expressão da CMáxEstFMEJcad-ext (absoluta, FMEJ/MC e FMEJ/MCb) foi empregada a regressão logística simples. Os pontos de corte para Sc a partir da CMáxEstFMEJcad-ext foram definidos pela curva Característica de Operação do Receptor (ROC) e localizados pelo índice de Youden. As análises foram realizadas no Statistical Product and Service Solutions (SPSS) 20.0 e MedCalc 15.2 com níveis de significância previamente estabelecidos (?=0,05). Os resultados indicaram que a Sc esteve presente em 12,9% dos homens e 9,0% das mulheres. Houve alta correlação entre a medida de FMEJ de referência e a Estimada (r=0,81), mesmo entre idosos com Sc (r=0,72). Os expoentes b obtidos foram de 0,96 e 0,70 para homens e mulheres, respectivamente. Na regressão logística, as expressões relativas (FMEJ/MC e FMEJ/MCb) não explicaram a probabilidade para ocorrência da Sc em nenhum dos sexos. Somente a CMáxEstFMEJcad-ext absoluta explicou a chance para homens (?2=3,869; p=0,049) e mulheres (?2=4,145; p=0,042). A área abaixo da curva foi elevada (AUC>0,70), com pontos de corte de 65,0kg para homens e 34,9kg para mulheres. Conclui-se que a CMáxEstFMEJcad-ext é uma medida válida para monitorar Sc como parâmetro de FM em idosos, mesmo quando apresentam Sc. Os limiares de carga (kg) propostos como pontos de corte podem ser usados em um simples teste FMEJ da prática clínica profissional. Além disso, tem boa sensibilidade para monitorar a distância do ponto corte para Sc, o que não é possível com o modelo dicotômico do EWGSOP / Sarcopenia (Sc) is a disease characterized by pathological symptoms without specific causes that affects part of the elderly. Sc promotes reductions in muscle mass (MM) and muscle strength (MS), with serious functional and motor impacts. The handgrip strength (HS) used to diagnose Sc is not representative of global MS, especially for lower limb (LL) MS, when the elderly go through an intervention with MS training. Although used to define Sc, MS is considered absolutely or relativized by body mass (BM), whose relationship is not always linear. Thus, the objective of this study was to establish valid parameters for the diagnosis of Sc, from different MS expressions of LL in the elderly. A sample of 98 physically independent elderly subjects was measured by dual energy absorptiometry (DXA) to determine appendicular lean soft tissue. The elderly were classified by sex and Sc (1=Sc; 0=not Sc), according to the criteria of the European Working Group on Sarcopenia in Older People (EWGSOP). The isokinetic knee extension MS (KEMS) determined at 60º/s (Biodex, System 4 Pro) was considered as referential (KEMSPeakTork-60°/s), and KEMS Estimated in extensor chair (EstMaxLoadKEMSext-ch), in maximal repetition protocol. Descriptive analysis with measures of central tendency was used to characterize the sample. Validation of the estimated KEMS was tested by correlation with the reference KEMS. The values of EstMaxLoadKEMSext-ch were relativized by body mass (KEMS/BM) and by allometric adjustments (KEMS/BMb), where b is the allometric exponent generated from the log-linear regression between KEMS and BM. To determine an explanatory model of Sc from each expression of EstMaxLoadKEMSext-ch (absolute, KEMS/BM and KEMS/BMb), simple logistic regression was used. The cutoff points for Sc from the EstMaxLoadKEMSext-ch were defined by the Receiver Operating Characteristic (ROC) curve and located by the Youden index. The analyzes were performed in Statistical Product and Service Solutions (SPSS) 20.0 and MedCalc 15.2 with previously established levels of significance (? = 0.05). The results indicated that Sc was present in 12.9% of men and 9.0% of women. There was a high correlation between the reference KEMS and the estimated (r=0.81), even among the elderly with Sc (r=0.72). The exponent b obtained was 0.96 and 0.70 for men and women, respectively. In the logistic regression, the relative expressions (KEMS/BM and KEMS/BMb) did not explain the probability for occurrence of Sc in any of the sexes. Only absolute EstMaxLoadKEMSext-ch explained the chance for males (?2=3,869, p=0.049) and females (?2=4.145, p=0.042). The area below the curve was elevated (AUC>0.70), with cutoff points of 65.0kg for men and 34.9kg for women. It is concluded that the EstMaxLoadKEMSext-ch is a valid measure to monitor Sc as MS parameter in the elderly, even when they present Sc. Load thresholds (kg) proposed as cutoff points can be used in a simple clinical practice test. In addition, it has good sensitivity to monitor the distance from the cut point to MS, which is not possible with the dichotomous model of the EWGSOP

1. Introducción y Objetivos: Existe un gran número de modelos predictivos de desenlaces en Insuficiencia Cardíaca Aguda; sin embargo, no se conoce la utilidad clínica. Se evaluará el desempeño de los modelos incluidos en el artículo. 2. Metodología: Se realizó una revisión sistemática de los modelos predictivos de desenlaces adversos. Por lo que se realizó una búsqueda en 6 bases bibliográficas, incluyéndose todo artículo original que reporte modelos de predicción. Se utilizó la herramienta PROBAST para valorar el riesgo de sesgo en los artículos. 3. Resultados: De 2498 artículos seleccionados, 39 fueron incluidos. La mayoría provienen de Europa (46.1%), 20 de ellos evalúan mortalidad, mientras que 21 demostraron bajo riesgo de sesgo y buena aplicabilidad utilizando la herramienta de evaluación PROBAST. El sexo masculino fue el mayoritario en el 69% de artículos. La morbilidad más común es hipertensión arterial. Todos los modelos fueron validados, y más del 50% cuentan con validación externa. Borovac JA et al. presenta el AUC más alto, con 0.907. Catorce modelos utilizaron el modelo de calibración de Hosmer-Lemeshow. Ningún artículo contó con beneficio neto. 4. Conclusión: : Los estudios no muestran una adecuada evaluación de desempeño, al no presentar todas las estadísticas de medición que existen; es necesario el desarrollo de nuevos modelos con la descripción completa de estadísticas de desempeño, para llevar a cabo una adecuada comparación. / 1. Introduction and Objectives: There are currently a number of predictive models of outcomes in Acute Heart Failure (AHF); the clinical usefulness and the possible implementation of these are unknown. In this study, we propose to evaluate the performance of the predictive models for AHF. 2. Methodology: A systematic review of the predictive models of adverse outcomes was performed. Therefore, a search was made in 6 bibliographic databases, including all original articles that report prediction models. The PROBAST tool was used to assess the risk of bias in the articles.3. Results: From 2498 initial articles, we took 39 from them. The majority came from Europe (46.1%), 20 of them assess mortality, while 21 demonstrated low risk of bias and good applicability using the PROBAST assessment tool. The male sex was the majority in 69% of articles. The most common morbidity is high blood pressure. All the models were validated, and more than 50% have external validation. Borovac JA et al. it has the highest AUC with 0.907. Fourteen models used the Hosmer Lemeshow calibration model. No article had a net benefit. 4. Conclusion: The studies don’t show an adequate evaluation of performance, since they do not present all the existing measurement statistics; The development of new models with the complete description of performance statistics is necessary to carry out an adequate comparison. / Tesis

El acceso a un sistema de salud digno constituye uno de los derechos fundamentales de toda persona, en el Perú se han realizado grandes esfuerzos para mejorar la calidad de los sistemas de salud, es un desafío al bicentenario el brindar un aseguramiento de salud de calidad que pueda alcanzar a todos los peruanos. Este objetivo enfrenta grandes desafíos ya que existen deficiencias en los procesos de las instituciones que brindan servicios de salud, siendo la Aseguradora Pública de Salud uno de los principales actores en el aseguramiento de la salud en el Perú. Nuestra investigación se ha centrado en el proceso de Evaluación Automática (EA), que tiene como objetivo evaluar la validez de las atenciones brindadas por las Instituciones Prestadoras de Servicios de Salud (IPRESS) afiliadas a la Aseguradora Pública de Salud. Durante los años 2017 y 2018 se detectó que el 3.82% y 1.85% del total de atenciones presentaban irregularidades. Estudios hechos a nivel mundial muestran que el nivel de irregularidades en entidades similares se encuentra entre el 3% y el 10% por lo que existe la posibilidad de elevar la capacidad de detección de irregularidades en la citada aseguradora. A través de nuestra investigación hemos identificado que mediante el uso de modelos predictivos construidos mediante la analítica de datos en el proceso de Evaluación Automática (EA), específicamente en la etapa llamada Supervisión Médica Electrónica (SME), se puede incrementar el nivel de detección de irregularidades, para ello es necesario aplicar la metodología CRISP-DM y el software WEKA. / Access to a decent health system is one of the fundamental rights of every person, in Peru great efforts have been made to improve the quality of health systems, it is a challenge for the bicentennial to provide quality health assurance that can reach all Peruvians. This objective faces great challenges since there are deficiencies in the processes of the institutions that provide health services, with the Public Health Insurer being one of the main actors in health insurance in Peru. Our research has focused on the Automatic Evaluation (EA) process, which aims to assess the validity of the health care provided by the Health Services Provider Institutions (IPRESS) affiliated with the Public Health Insurer. During the years 2017 and 2018 it was detected that 3.82% and 1.85% of the total health care presented irregularities. Studies done worldwide show that the level of irregularities in similar entities is between 3% and 10%, so there is the possibility of increasing the ability to detect irregularities in the aforementioned insurer. Through our research we have identified that by using predictive models constructed through data analytics in the Automatic Evaluation (EA) process, specifically at the stage called Electronic Medical Supervision (SME), it is posible to increase the level of irregularity detection, for this it is necessary to apply the CRISP-DM methodology and the WEKA software. / Trabajo de investigación

Stress fractures are common in the limb bones of human and equine athletes alike. Repetitive skeletal loading can lead to remodeling and the accumulation of microdamage in bone, which only becomes grossly evident during catastrophic fracture of the bone due to the accumulated microdamage. Though various metrics attempting to quantify bone health exist, none have distinguished themselves as early predictors of the susceptibility of bone to fracture. In this exploratory study, we examine the ability of several evaluation methods to distinguish between third metacarpal (MC3) bones from racehorses that have experienced a limb-bone fracture and from those that have not. Third metacarpal bones were harvested from deceased Thoroughbred racehorses and categorized into four groups: MC3 bones from horses whose cause of death was not related to skeletal fracture (Control group, n = 20), MC3 bones form horses that were euthanized after fracturing proximal sesamoid bones (Sesamoid group, n = 20), MC3 bones from horses that were euthanized after fracturing a non-MC3 long bone (Long Bone group, n = 19), and MC3 bones from horses that were euthanized after fracturing an MC3 (MC3 group, n = 5). Each MC3 bone underwent testing using a variety of tools and methods at the proximal, midshaft, and distal levels of the lateral, dorsal, and medial surfaces. All tools and methods (OsteoProbe reference point indentation, BioDent reference point indentation, x-ray, micro-CT, and pQCT) exhibited some capability in differentiating between control and fracture groups. The long-term objective of this project is to create a model that will utilize data from a set of evaluations and output the susceptibility of the horse to fracture a bone, a long bone, or the MC3, specifically. Although the sample size in this study is not sufficient to create a reliably predictive logistic regression model, promising results from preliminary models provide incentive to further explore the possibility of creating one. While clinical practicality will be a vital consideration for a model in the future, establishing this basis for the capability of each evaluation at hand is a necessary first step in predicting and preventing fracture in bone.

Sepsis is an organ dysfunction life-threatening disease that is caused by a dysregulated body response to infection. Sepsis is difficult to detect at an early stage, and when not detected early, is difficult to treat and results in high mortality rates. Developing improved methods for identifying patients in high risk of suffering septic shock has been the focus of much research in recent years. Building on this body of literature, this dissertation develops an improved method for septic shock prediction. Using the data from the MMIC-III database, an ensemble classifier is trained to identify high-risk patients. A robust prediction model is built by obtaining a risk score from fitting the Cox Hazard model on multiple input features. The score is added to the list of features and the Random Forest ensemble classifier is trained to produce the model. The Cox Enhanced Random Forest (CERF) proposed method is evaluated by comparing its predictive accuracy to those of extant methods.

京都大学 / 0048 / 新制・課程博士 / 博士(社会健康医学) / 甲第18191号 / 社医博第52号 / 新制||社医||8(附属図書館) / 31049 / 京都大学大学院医学研究科社会健康医学系専攻 / (主査)教授 中山 健夫, 教授 佐藤 俊哉, 教授 木村 剛 / 学位規則第4条第1項該当 / Doctor of Public Health / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第19984号 / 工博第4228号 / 新制||工||1654(附属図書館) / 33080 / 京都大学大学院工学研究科都市環境工学専攻 / (主査)教授 田中 宏明, 教授 米田 稔, 講師 山下 尚之 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

Predictive models based in symptomatology of suspected dengue patients seeking medical care in Honduras. The models based on logistic regression models predicted the outcomes of dengue fever/ severe dengue. Sensitivity and specificity are discussed. It also describe the level of agreement between Honduran classification of severe dengue and the ones based on World Health Organization guidelines of 1997 and 2009. / Introduction: Dengue is a major public health problem in tropical and subtropical countries but its clinical presentation may be similar to many febrile illnesses. Since in endemic countries laboratory confirmation is frequently delayed, the majority of dengue cases are diagnosed based on patient’s symptomatology. This can often lead to misdiagnosis and potential serious health complications. The objective of this study was to identify clinical, hematological and demographical parameters that could be used as predictors of dengue fever among patients with febrile illness. Methods: We conducted a retrospective cohort study of 548 patients presenting with febrile syndrome to the largest public hospitals in Honduras. Patients’ clinical, laboratory, and demographical data as well as dengue laboratory confirmation by either serology or viral isolation were used to build a predictive statistical model to identify dengue cases. Results: Of 548 patients, 390 were confirmed with dengue infection while 158 had negative results. Univariable analysis revealed seven variables associated with dengue: male sex, petechiae, skin rash, myalgia, retro-ocular pain, positive tourniquet test, and bleeding gums. In multivariable logistic regression analysis, retro-ocular pain petechiae and bleeding gums were associated with increased risk, while epistaxis and paleness of skin were associated with reduced risk of dengue. Using a value of 0.6 (i.e., 60% probability for a case to be positive based on the equation values), our model had a sensitivity of 86.2%, a specificity of 27.2%, and an overall accuracy of 69.2%; allowing for the diagnosis of dengue to be ruled out and for other febrile conditions to be investigated. Conclusions: The application of predictive models can be valuable when laboratory confirmation is delayed. Among Honduran patients presenting with febrile illness, our data reveal key symptoms associated with dengue fever, however the overall accuracy of our model is still low and specificity remains a concern. Our model requires validation in other populations with similar pattern of dengue transmission. Key Words: Dengue, fever, Predictive model, symptoms, Honduras / Thesis / Doctor of Philosophy (PhD) / Predictive models based in symptomatology of suspected dengue patients seeking medical care in Honduras. The models based on logistic regression models predicted the outcomes of dengue fever/ severe dengue. Sensitivity and specificity are discussed. It also describe the level of agreement between Honduran classification of severe dengue and the ones based on World Health Organization guidelines of 1997 and 2009.

In recent years, the scope of Machine Learning algorithms and its techniques are taking up a notch in every industry (for example, recommendation systems, user behavior analytics, financial applications and many more). In practice, they play an important role in utilizing the power of the vast data we currently generate on a daily basis in our digital world.In this study, we present a comprehensive comparison of different supervised Machine Learning algorithms and feature selection techniques to build a best predictive model as an output. Thus, this predictive model helps companies predict unwanted behavior in their business processes. In addition, we have researched for the automation of all the steps involved (from understanding data to implementing models) in the complete Machine Learning Pipeline, also known as AutoML, and provide a comprehensive survey of the various frameworks introduced in this domain. These frameworks were introduced to solve the problem of CASH (combined algorithm selection and Hyper- parameter optimization), which is basically automation of various pipelines involved in the process of building a Machine Learning predictive model. / Under de senaste åren har omfattningen av maskininlärnings algoritmer och tekniker tagit ett steg i alla branscher (till exempel rekommendationssystem, beteendeanalyser av användare, finansiella applikationer och många fler). I praktiken spelar de en viktig roll för att utnyttja kraften av den enorma mängd data vi för närvarande genererar dagligen i vår digitala värld.I den här studien presenterar vi en omfattande jämförelse av olika övervakade maskininlärnings algoritmer och funktionsvalstekniker för att bygga en bästa förutsägbar modell som en utgång. Således hjälper denna förutsägbara modell företag att förutsäga oönskat beteende i sina affärsprocesser. Dessutom har vi undersökt automatiseringen av alla inblandade steg (från att förstå data till implementeringsmodeller) i den fullständiga maskininlärning rörledningen, även känd som AutoML, och tillhandahåller en omfattande undersökning av de olika ramarna som introducerats i denna domän. Dessa ramar introducerades för att lösa problemet med CASH (kombinerat algoritmval och optimering av Hyper-parameter), vilket i grunden är automatisering av olika rörledningar som är inblandade i processen att bygga en förutsägbar modell för maskininlärning.