Multiple testing in spatial epidemiology: a Bayesian approach

In this work we aim to propose a new approach for preliminary epidemiological studies on Standardized Mortality Ratios (SMR) collected in many spatial regions. A preliminary study on SMRs aims to formulate hypotheses to be investigated via individual epidemiological studies that avoid bias carried on by aggregated analyses. Starting from collecting disease counts and calculating expected disease counts by means of reference population disease rates, in each area an SMR is derived as the MLE under the Poisson assumption on each observation. Such estimators have high standard errors in small areas, i.e. where the expected count is low either because of the low population underlying the area or the rarity of the disease under study. Disease mapping models and other techniques for screening disease rates among the map aiming to detect anomalies and possible high-risk areas have been proposed in literature according to the classic and the Bayesian paradigm. Our proposal is approaching this issue by a decision-oriented method, which focus on multiple testing control, without however leaving the preliminary study perspective that an analysis on SMR indicators is asked to. We implement the control of the FDR, a quantity largely used to address multiple comparisons problems in the eld of microarray data analysis but which is not usually employed in disease mapping. Controlling the FDR means providing an estimate of the FDR for a set of rejected null hypotheses. The small areas issue arises diculties in applying traditional methods for FDR estimation, that are usually based only on the p-values knowledge (Benjamini and Hochberg, 1995; Storey, 2003). Tests evaluated by a traditional p-value provide weak power in small areas, where the expected number of disease cases is small. Moreover tests cannot be assumed as independent when spatial correlation between SMRs is expected, neither they are identical distributed when population underlying the map is heterogeneous. The Bayesian paradigm oers a way to overcome the inappropriateness of p-values based methods. Another peculiarity of the present work is to propose a hierarchical full Bayesian model for FDR estimation in testing many null hypothesis of absence of risk.We will use concepts of Bayesian models for disease mapping, referring in particular to the Besag York and Mollié model (1991) often used in practice for its exible prior assumption on the risks distribution across regions. The borrowing of strength between prior and likelihood typical of a hierarchical Bayesian model takes the advantage of evaluating a singular test (i.e. a test in a singular area) by means of all observations in the map under study, rather than just by means of the singular observation. This allows to improve the power test in small areas and addressing more appropriately the spatial correlation issue that suggests that relative risks are closer in spatially contiguous regions. The proposed model aims to estimate the FDR by means of the MCMC estimated posterior probabilities b i's of the null hypothesis (absence of risk) for each area. An estimate of the expected FDR conditional on data (\FDR) can be calculated in any set of b i's relative to areas declared at high-risk (where thenull hypothesis is rejected) by averaging the b i's themselves. The\FDR can be used to provide an easy decision rule for selecting high-risk areas, i.e. selecting as many as possible areas such that the\FDR is non-lower than a prexed value; we call them\FDR based decision (or selection) rules. The sensitivity and specicity of such rule depend on the accuracy of the FDR estimate, the over-estimation of FDR causing a loss of power and the under-estimation of FDR producing a loss of specicity. Moreover, our model has the interesting feature of still being able to provide an estimate of relative risk values as in the Besag York and Mollié model (1991). A simulation study to evaluate the model performance in FDR estimation accuracy, sensitivity and specificity of the decision rule, and goodness of estimation of relative risks, was set up. We chose a real map from which we generated several spatial scenarios whose counts of disease vary according to the spatial correlation degree, the size areas, the number of areas where the null hypothesis is true and the risk level in the latter areas. In summarizing simulation results we will always consider the FDR estimation in sets constituted by all b i's selected lower than a threshold t. We will show graphs of the\FDR and the true FDR (known by simulation) plotted against a threshold t to assess the FDR estimation. Varying the threshold we can learn which FDR values can be accurately estimated by the practitioner willing to apply the model (by the closeness between\FDR and true FDR). By plotting the calculated sensitivity and specicity (both known by simulation) vs the\FDR we can check the sensitivity and specicity of the corresponding\FDR based decision rules. For investigating the over-smoothing level of relative risk estimates we will compare box-plots of such estimates in high-risk areas (known by simulation), obtained by both our model and the classic Besag York Mollié model. All the summary tools are worked out for all simulated scenarios (in total 54 scenarios). Results show that FDR is well estimated (in the worst case we get an overestimation, hence a conservative FDR control) in small areas, low risk levels and spatially correlated risks scenarios, that are our primary aims. In such scenarios we have good estimates of the FDR for all values less or equal than 0.10. The sensitivity of\FDR based decision rules is generally low but specicity is high. In such scenario the use of\FDR = 0:05 or\FDR = 0:10 based selection rule can be suggested. In cases where the number of true alternative hypotheses (number of true high-risk areas) is small, also FDR = 0:15 values are well estimated, and \FDR = 0:15 based decision rules gains power maintaining an high specicity. On the other hand, in non-small areas and non-small risk level scenarios the FDR is under-estimated unless for very small values of it (much lower than 0.05); this resulting in a loss of specicity of a\FDR = 0:05 based decision rule. In such scenario\FDR = 0:05 or, even worse,\FDR = 0:1 based decision rules cannot be suggested because the true FDR is actually much higher. As regards the relative risk estimation, our model achieves almost the same results of the classic Besag York Molliè model. For this reason, our model is interesting for its ability to perform both the estimation of relative risk values and the FDR control, except for non-small areas and large risk level scenarios. A case of study is nally presented to show how the method can be used in epidemiology.

Tecniche MonteCarlo per l'analisi di sistemi a rete

The assessment of the RAMS (Reliability, Availability, Maintainability and Safety) performances of system generally includes the evaluations of the “Importance” of its components and/or of the basic parameters of the model through the use of the Importance Measures. The analytical equations proposed in this study allow the estimation of the first order Differential Importance Measure on the basis of the Birnbaum measures of components, under the hypothesis of uniform percentage changes of parameters. The aging phenomena are introduced into the model by assuming exponential-linear or Weibull distributions for the failure probabilities. An algorithm based on a combination of MonteCarlo simulation and Cellular Automata is applied in order to evaluate the performance of a networked system, made up of source nodes, user nodes and directed edges subjected to failure and repair. Importance Sampling techniques are used for the estimation of the first and total order Differential Importance Measures through only one simulation of the system “operational life”. All the output variables are computed contemporaneously on the basis of the same sequence of the involved components, event types (failure or repair) and transition times. The failure/repair probabilities are forced to be the same for all components; the transition times are sampled from the unbiased probability distributions or it can be also forced, for instance, by assuring the occurrence of at least a failure within the system operational life. The algorithm allows considering different types of maintenance actions: corrective maintenance that can be performed either immediately upon the component failure or upon finding that the component has failed for hidden failures that are not detected until an inspection; and preventive maintenance, that can be performed upon a fixed interval. It is possible to use a restoration factor to determine the age of the component after a repair or any other maintenance action.

Analisi critica di modelli previsionali per le frane in Emilia Romagna

Questa tesi di dottorato è inserita nell’ambito della convenzione tra ARPA_SIMC (che è l’Ente finanziatore), l’Agenzia Regionale di Protezione Civile ed il Dipartimento di Scienze della Terra e Geologico - Ambientali dell’Ateneo di Bologna. L’obiettivo principale è la determinazione di possibili soglie pluviometriche di innesco per i fenomeni franosi in Emilia Romagna che possano essere utilizzate come strumento di supporto previsionale in sala operativa di Protezione Civile. In un contesto geologico così complesso, un approccio empirico tradizionale non è sufficiente per discriminare in modo univoco tra eventi meteo innescanti e non, ed in generale la distribuzione dei dati appare troppo dispersa per poter tracciare una soglia statisticamente significativa. È stato quindi deciso di applicare il rigoroso approccio statistico Bayesiano, innovativo poiché calcola la probabilità di frana dato un certo evento di pioggia (P(A|B)) , considerando non solo le precipitazioni innescanti frane (quindi la probabilità condizionata di avere un certo evento di precipitazione data l’occorrenza di frana, P(B|A)), ma anche le precipitazioni non innescanti (quindi la probabilità a priori di un evento di pioggia, P(A)). L’approccio Bayesiano è stato applicato all’intervallo temporale compreso tra il 1939 ed il 2009. Le isolinee di probabilità ottenute minimizzano i falsi allarmi e sono facilmente implementabili in un sistema di allertamento regionale, ma possono presentare limiti previsionali per fenomeni non rappresentati nel dataset storico o che avvengono in condizioni anomale. Ne sono esempio le frane superficiali con evoluzione in debris flows, estremamente rare negli ultimi 70 anni, ma con frequenza recentemente in aumento. Si è cercato di affrontare questo problema testando la variabilità previsionale di alcuni modelli fisicamente basati appositamente sviluppati a questo scopo, tra cui X – SLIP (Montrasio et al., 1998), SHALSTAB (SHALlow STABility model, Montgomery & Dietrich, 1994), Iverson (2000), TRIGRS 1.0 (Baum et al., 2002), TRIGRS 2.0 (Baum et al., 2008).

Earthquake forecasting and seismic hazard analysis: some insights on the testing phase and the modeling

This thesis is divided in three chapters. In the first chapter we analyse the results of the world forecasting experiment run by the Collaboratory for the Study of Earthquake Predictability (CSEP). We take the opportunity of this experiment to contribute to the definition of a more robust and reliable statistical procedure to evaluate earthquake forecasting models. We first present the models and the target earthquakes to be forecast. Then we explain the consistency and comparison tests that are used in CSEP experiments to evaluate the performance of the models. Introducing a methodology to create ensemble forecasting models, we show that models, when properly combined, are almost always better performing that any single model. In the second chapter we discuss in depth one of the basic features of PSHA: the declustering of the seismicity rates. We first introduce the Cornell-McGuire method for PSHA and we present the different motivations that stand behind the need of declustering seismic catalogs. Using a theorem of the modern probability (Le Cam's theorem) we show that the declustering is not necessary to obtain a Poissonian behaviour of the exceedances that is usually considered fundamental to transform exceedance rates in exceedance probabilities in the PSHA framework. We present a method to correct PSHA for declustering, building a more realistic PSHA. In the last chapter we explore the methods that are commonly used to take into account the epistemic uncertainty in PSHA. The most widely used method is the logic tree that stands at the basis of the most advanced seismic hazard maps. We illustrate the probabilistic structure of the logic tree, and then we show that this structure is not adequate to describe the epistemic uncertainty. We then propose a new probabilistic framework based on the ensemble modelling that properly accounts for epistemic uncertainties in PSHA.

The Impact Of Inward Licensing On New Venture’s Performance. Is inward licensing a winning strategy?

The original idea of the thesis draws on interrelated assumptions: 1) among the tools used, in the markets for technology, for the acquisition of external knowledge, the licensing agreements are acknowledged as one of the most important contractual mechanisms; 2) the liabilities of newness and the liabilities of smallness force new venture to strongly rely on external knowledge sources. Albeit the relevance of this topic, little attention has been paid so far to its investigation, especially in the licensing context; 3) nowadays there is an increasing trend in licensing practices, but the literature on markets for technology focuses almost exclusively on the incentives and rationales that foster firms’ decisions to trade their technologies, under-investigating the role of the acquiring firm, the licensee, overlooking the demand side of the market. Therefore, the thesis investigates the inward licensing phenomenon within the context of new ventures. The main questions that new venture licensee has to address if it decides to undertake an inward licensing strategy, can be summarized as follows: 1) Is convenient for a new venture to choose, as initial technology strategy, the implementation of an inward licensing ? 2) Does this decision affect its survival probabilities? 3) Does the age, at which a new venture becomes a licensee, affect its innovative capabilities? Is it better to undertake a licensing-in strategy soon after founding or to postpone this strategy until the new venture has accumulated significant resources? The findings suggest that new ventures licensees survive less than their non-licensee counterparts; the survival rates are directly connected to the time taken by firms to reach the market;being engaged in licensing-in deals some years after its inception allows a new venture licensee to increase its subsequent capacity to produce innovations.

Permutation-based stochastic ordering using pairwise comparisons

The topic of this work concerns nonparametric permutation-based methods aiming to find a ranking (stochastic ordering) of a given set of groups (populations), gathering together information from multiple variables under more than one experimental designs. The problem of ranking populations arises in several fields of science from the need of comparing G>2 given groups or treatments when the main goal is to find an order while taking into account several aspects. As it can be imagined, this problem is not only of theoretical interest but it also has a recognised relevance in several fields, such as industrial experiments or behavioural sciences, and this is reflected by the vast literature on the topic, although sometimes the problem is associated with different keywords such as: "stochastic ordering", "ranking", "construction of composite indices" etc., or even "ranking probabilities" outside of the strictly-speaking statistical literature. The properties of the proposed method are empirically evaluated by means of an extensive simulation study, where several aspects of interest are let to vary within a reasonable practical range. These aspects comprise: sample size, number of variables, number of groups, and distribution of noise/error. The flexibility of the approach lies mainly in the several available choices for the test-statistic and in the different types of experimental design that can be analysed. This render the method able to be tailored to the specific problem and the to nature of the data at hand. To perform the analyses an R package called SOUP (Stochastic Ordering Using Permutations) has been written and it is available on CRAN.