A raiva dos bovinos permanece endêmica no Brasil e apesar dos esforços no controle, a doença ainda se espalha insidiosamente. O principal transmissor é o morcego hematófago Desmodus rotundus. O presente trabalho objetivou o desenvolvimento de um modelo preditivo qualitativo para a ocorrência da raiva de bovinos, por município, em 25 dos 27 Estados brasileiros. O risco de transmissão de raiva de morcegos hematófagos para bovinos foi estimado utilizando modelos baseados em árvores decisórias de receptividade e vulnerabilidade. Questionários abrangendo questões relacionadas à vigilância de possíveis fatores de risco, como focos em bovinos, presença de abrigos de morcegos, morcegos positivos para ao vírus da raiva e mudanças ambientais foram aplicados às unidades locais veterinárias de cada Estado. A densidade de bovinos e as características geomorfológicas foram obtidas através de bases de dados nacionais e sistemas de informação geográfica. Dos 433 municípios que apresentaram focos de raiva bovina em 2010, 178 (41.1%) foram classificados pelo modelo como de alto risco, 212 (49.0%) foram classificados como de risco moderado, 25 (5.8%) com baixo risco, enquanto o risco não pode ser determinado em 18 (4.1%) municípios. Uma curva ROC foi desenvolvida para determinar se o risco avaliado pelo modelo poderia discriminar adequadamente os municípios em relação à ocorrência de raiva nos anos seguintes. O estimador de risco para os anos de 2011 e 2012 foi classificado como moderadamente acurado. No futuro, estes modelos poderão permitir o direcionamento de esforços no controle da raiva, com a adoção de medidas de controle direcionadas para áreas de maior risco e a otimização do deslocamento das equipes veterinários de campo pelo território nacional. Adicionalmente, esforços deve ser realizados para encorajar uma vigilância contínua dos fatores de risco. / Bovine rabies remains endemic in Brazil and despite control efforts, the disease still spreads insidiously. The main vector is the hematophagous bat, Desmodus rotundus. The present work aimed to create a predictive qualitative model of the occurrence of bovine rabies in each municipality in 25 of the 27 Brazilian States. The risk of rabies transmission from bats to bovine was estimated using decision-tree models of receptivity and vulnerability. Questionnaires, which covered a number of questions related to the surveillance of possible risk factors, such as bovine rabies outbreaks in the previous year, presence of bat roosts, bat rabies positivity and environmental changes, were sent to the local veterinary units of each State. The bovine density and geomorphologic features were obtained from national databases and geographic information systems. Of the 433 municipalities presenting bovine rabies outbreaks in 2010, 178 (41.1%) were classified by the model as high risk, 212 (49.0%) were classified as moderate risk, 25 (5.8%) were classified as low risk, whereas the risk was undetermined in 18 municipalities (4.1%). An ROC curve was built to determine if the risk evaluated by the model could adequately discriminate between municipalities with and without rabies occurrence in future years. The risk estimator for the year 2011 was classified as moderately accurate. In the future, these models could allow the targeting of rabies control efforts, with the adoption of control measures directed to the higher risk locations and the optimization of the field veterinary staff deployment throughout the country. Additionally, efforts must be made to encourage continuous surveillance of risk factors.

This research combines relational database framework and quantitative equity portfolio models based on the Barra Risk Model Handbook standard steps to design a database and computer platform for multi-factor risk management tasks. The multi-factor model facilitates fast search and efficient selection of descriptors with explanatory power for future stock returns. The design of database is divided into three steps. First, descriptors are calculated and daily-update modules constructed. This study finds 48 key descriptors which play important roles in explaining stock returns of Taiwan. Second, entity relational model is applied to sort out linkages between pieces of important information in the factor model. Lastly, database auto-run procedures are setup to update the latest raw data on a monthly basis. Model parameter update and portfolio rebalancing is hence made seamless to meet practical operation demand for such a platform. The development of the Multi-factor risk model is divided into five main steps. (1) Finding significant descriptors. (2) Forming common factors from descriptors. (3) Developing a multi-factor return model. (4) Developing a multi-factor risk model. (5) Running performance analysis and back-testing. The empirical results show that the average adjusted R-squared of the MFM model is 0.5 during the period of 1998/04~2005/11. For combining descriptors into common factors, we run factor analysis. The multi-collinearity problem existing in the descriptors is well taken care of by such procedures. We use the exponentially weighted averaging method to compute the factor returns and forecast stock ranking. A half-life of 24 months appears to deliver the best performance in Taiwan stock market.

This dissertation consists of three essays on credit risk models and their Bayesian estimation. In each essay, defaults or default correlation models are built under one of two main streams. In our first essay, sequential estimation on hidden asset value and model parameters estimation are implemented under the Black-Cox model. To capture short-term autocorrelation in the stock market, we assume that market noise follows a mean reverting process. For estimation, two Bayesian methods are applied in this essay: the particle filter algorithm for sequential estimation of asset value and the generalized Gibbs sampling method for model parameters estimation. The first simulation study shows that sequential hidden asset value estimation using option price and equity price is more efficient and accurate than estimation using only equity price. The second simulation study shows that by applying the generalized Gibbs sampling method, model parameters can be successfully estimated under the model setting that there is no closed-form solution. In an empirical analysis using eight companies, half of which are DowJones30 companies and the other half non-Dow Jones 30 companies, the stock market noise for the firms with more liquid stock is estimated as having smaller volatility in market noise processes. In our second essay, the frailty idea described in Duffie, Eckner, Horel, and Saita (2009) is expanded to industry-specific terms. The MCEM algorithm is used to estimate parameters and random effect processes under the condition of unknown hidden paths and analytically-difficult likelihood functions. The estimate used in the study are based on U.S. public firms between 1990 and 2008. By introducing industry-specific hidden factors and assuming that they are random effects, a comparison is made of the relative scale of within- and between-industries correlations. A comparison study is also developed among a without-hidden-factor model, a common-hiddenfactor model, and our industry-specific common-factor model. The empirical results show that an industry-specific common factor is necessary for adjusting over- or under-estimation of default probabilities and over- or under-estimation of observed common factor effects. Our third essay combines and extends works of the first two essays by proposing a common model frame for both structural and intensity credit risk models. The common model frame combines the merits of several default correlation studies which are independently developed under each model setting. Following the work of Duffie, Eckner, Horel, and Saita (2009), we apply not only observed common factors, but also un-observed hidden factor to explain the correlated defaults. Bayesian techniques are used for estimation and generalized Gibbs sampling and Metropolis-Hasting (MH) algorithms are developed. More than a simple combination of two model approaches (structural and intensity models), we relax the assumptions of equal factor effect across entire firms in previous studies, instead adopting a random coefficients model. Also, a novelty of the approach lies in the fact that CDS and equity prices are used together for estimation. A simulation study shows that the posterior convergence is improved by adding CDS prices in estimation. Empirical results based on daily data of 125 companies comprising CDS.NA.IG13 in 2009 supports the necessity of such relaxations of assumption in previous studies. In order to demonstrate potential practical applications of the proposed framework, we derive the posterior distribution of CDX tranche prices. Our correlated structural model is successfully able to predict all the CDX tranche prices, but our correlated intensity model results suggests the need for further modification of the model. / Statistics

Introduktion Kritik har riktats mot utformningen av revisionsriskmodellenoch trots det är den en viktig del förbedömningen av revisionsrisken i planeringsfasen. Ioch med detta är det relevant att studera användarnas,det vill säga revisorernas, syn på revisionsriskmodellenoch dess användning. Syfte Uppsatsens syfte är att förklara revisorers uppfattningom användningen av de olika riskerna i revisionsriskmodellen. Metod Studien använder sig av en kvantitativ metod. Vigenomför en enkätundersökning med auktoriseraderevisorer. Data samlas in med hjälp av SurveyMonkeyoch analyseras med SPSS. Slutsatser Våra resultat visar att revisorer som medverkat iundersökningen upplever att inneboende risk varviktigast följt av kontrollrisk och därefter upptäcktsrisk.Vi ser dock att bedömningen av inneboende riskoch kontrollrisk sker sammanvägt och att de påverkasav varandra vilket stärker ett flertal tidigare studier. / Introduction Criticism has been leveled at the design of theaudit risk model, even though it is an importantpart of the assessment of audit risk in the planningphase. As a result of this, it is relevant to studyusers’, namely auditors’, approach to the audit riskmodel and its use. Purpose The purpose of this thesis is to explain auditors'interpretation of the use of the various risks in theaudit risk model. Method The study uses a quantitative approach. The datahas been collected through a survey usingSurveyMonkey and then analyzed with SPSS. Conclusions Our results show that the auditors who participatedin the survey felt that inherent risk was mostimportant, followed by control risk, and thereafterdetection risk. However, we see that theassessment of inherent and control risks iscombined and that they are influenced by eachother, which strengthens several previous studies.

In this thesis, we consider the problem of constructing an additive risk model based on the right censored survival data to predict the survival times of the cancer patients, especially when the dimension of the covariates is much larger than the sample size. For microarray Gene Expression data, the number of gene expression levels is far greater than the number of samples. Such ¡°small n, large p¡± problems have attracted researchers to investigate the association between cancer patient survival times and gene expression profiles for recent few years. We apply Partial Least Squares to reduce the dimension of the covariates and get the corresponding latent variables (components), and these components are used as new regressors to fit the extensional additive risk model. Also we employ the time dependent AUC curve (area under the Receiver Operating Characteristic (ROC) curve) to assess how well the model predicts the survival time. Finally, this approach is illustrated by re-analysis of the well known AML data set and breast cancer data set. The results show that the model fits both of the data sets very well.

It is of interest that researchers study competing risks in which subjects may fail from any one of K causes. Comparing any two competing risks with covariate effects is very important in medical studies. This thesis develops omnibus tests for comparing cause-specific hazard rates and cumulative incidence functions at specified covariate levels. In the thesis, the omnibus tests are derived under the additive risk model, that is an alternative to the proportional hazard model, with by a weighted difference of estimates of cumulative cause-specific hazard rates. Simultaneous confidence bands for the difference of two conditional cumulative incidence functions are also constructed. A simulation procedure is used to sample from the null distribution of the test process in which the graphical and numerical techniques are used to detect the significant difference in the risks. A melanoma data set is used for the purpose of illustration.

One problem of interest is to relate genes to survival outcomes of patients for the purpose of building regression models to predict future patients¡¯ survival based on their gene expression data. Applying semeparametric additive risk model of survival analysis, this thesis proposes a new approach to conduct the analysis of gene expression data with the focus on model¡¯s predictive ability. The method modifies the correlation principal component regression to handle the censoring problem of survival data. Also, we employ the time dependent AUC and RMSEP to assess how well the model predicts the survival time. Furthermore, the proposed method is able to identify significant genes which are related to the disease. Finally, this proposed approach is illustrated by simulation data set, the diffuse large B-cell lymphoma (DLBCL) data set, and breast cancer data set. The results show that the model fits both of the data sets very well.

This paper presents an expert system to monitor social sustainability compliance in supply chains. The system allows to continuously rank suppliers based on their risk of breaching sustainability standards on child labor. It uses a Bayesian network to determine the breach likelihood for each supplier location based on the integration of statistical data, audit results and public reports of child labor incidents. Publicly available statistics on the frequency of child labor in different regions and industries are used as contextual prior. The impact of audit results on the breach likelihood is calibrated based on expert input. Child labor incident observations are included automatically from publicly available news sources using text mining algorithms. The impact of an observation on the breach likelihood is determined by its relevance, credibility and frequency. Extensive tests reveal that the expert system correctly replicates the decisions of domain experts in the fields supply chain management, sustainability management, and risk management.

Infectious disease dynamics are inherently shaped by the distribution, ecology, and genetic variation of hosts. Conversely, pathogens exert powerful influences on hosts through demographic processes and natural selection. These tenets of disease ecology and evolutionary biology are illustrated in the case of white-nose syndrome (WNS), an emerging infectious disease of hibernating bats. WNS first emerged in 2006 and spread rapidly throughout eastern North America, causing massive declines in bat populations. To understand how host ecology and spatial distribution influence the spread of WNS, I evaluated risk models of colony-level correlates, including bat colony size, species composition, behavior, and gene flow. WNS was more likely to emerge in large colonies first, and species composition and behavior were also significant predictors of risk. Spatial spread was predicted by population genetics of little brown myotis (Myotis lucifugus), indicating coupling of host gene flow and pathogen dispersal, and potential for the application of landscape genetics to predict future spread. To guide management and evaluate pre-existing genetic diversity, I assessed population genetic structure of little brown myotis using restriction site-associated DNA sequencing (RAD-seq). RAD-seq data revealed two populations divided by the Rocky Mountains, with high gene flow between the distributions of putative subspecies. Demographic analyses and genome scans suggest adaptive genetic variation, variation that may be threatened by WNS in eastern North America. Drastic declines from WNS have likely imposed strong selection, and recent stabilization of populations near the disease epicenter suggests that resistance may have evolved in the host population. I generated whole genome sequence data for bats sampled before and after declines to test for demographic changes and natural selection. Average genomic differentiation and nucleotide diversity indicated little demographic change between the two periods, but preliminary analyses suggest genomic regions of differentiation combined with decreased nucleotide diversity in post-WNS relative to pre-WNS samples, hinting at a pattern of natural selection. Additional samples and in-depth analyses are necessary to robustly test these patterns; however, identification of signatures of selection in the bat genome would be an exciting indication of a rapid evolutionary response to an introduced disease.

A raiva dos bovinos permanece endêmica no Brasil e apesar dos esforços no controle, a doença ainda se espalha insidiosamente. O principal transmissor é o morcego hematófago Desmodus rotundus. O presente trabalho objetivou o desenvolvimento de um modelo preditivo qualitativo para a ocorrência da raiva de bovinos, por município, em 25 dos 27 Estados brasileiros. O risco de transmissão de raiva de morcegos hematófagos para bovinos foi estimado utilizando modelos baseados em árvores decisórias de receptividade e vulnerabilidade. Questionários abrangendo questões relacionadas à vigilância de possíveis fatores de risco, como focos em bovinos, presença de abrigos de morcegos, morcegos positivos para ao vírus da raiva e mudanças ambientais foram aplicados às unidades locais veterinárias de cada Estado. A densidade de bovinos e as características geomorfológicas foram obtidas através de bases de dados nacionais e sistemas de informação geográfica. Dos 433 municípios que apresentaram focos de raiva bovina em 2010, 178 (41.1%) foram classificados pelo modelo como de alto risco, 212 (49.0%) foram classificados como de risco moderado, 25 (5.8%) com baixo risco, enquanto o risco não pode ser determinado em 18 (4.1%) municípios. Uma curva ROC foi desenvolvida para determinar se o risco avaliado pelo modelo poderia discriminar adequadamente os municípios em relação à ocorrência de raiva nos anos seguintes. O estimador de risco para os anos de 2011 e 2012 foi classificado como moderadamente acurado. No futuro, estes modelos poderão permitir o direcionamento de esforços no controle da raiva, com a adoção de medidas de controle direcionadas para áreas de maior risco e a otimização do deslocamento das equipes veterinários de campo pelo território nacional. Adicionalmente, esforços deve ser realizados para encorajar uma vigilância contínua dos fatores de risco. / Bovine rabies remains endemic in Brazil and despite control efforts, the disease still spreads insidiously. The main vector is the hematophagous bat, Desmodus rotundus. The present work aimed to create a predictive qualitative model of the occurrence of bovine rabies in each municipality in 25 of the 27 Brazilian States. The risk of rabies transmission from bats to bovine was estimated using decision-tree models of receptivity and vulnerability. Questionnaires, which covered a number of questions related to the surveillance of possible risk factors, such as bovine rabies outbreaks in the previous year, presence of bat roosts, bat rabies positivity and environmental changes, were sent to the local veterinary units of each State. The bovine density and geomorphologic features were obtained from national databases and geographic information systems. Of the 433 municipalities presenting bovine rabies outbreaks in 2010, 178 (41.1%) were classified by the model as high risk, 212 (49.0%) were classified as moderate risk, 25 (5.8%) were classified as low risk, whereas the risk was undetermined in 18 municipalities (4.1%). An ROC curve was built to determine if the risk evaluated by the model could adequately discriminate between municipalities with and without rabies occurrence in future years. The risk estimator for the year 2011 was classified as moderately accurate. In the future, these models could allow the targeting of rabies control efforts, with the adoption of control measures directed to the higher risk locations and the optimization of the field veterinary staff deployment throughout the country. Additionally, efforts must be made to encourage continuous surveillance of risk factors.