Cette étude est née de l’observation d’un décalage entre d’une part, l’êthos que l’on fabrique à Plutarque à la Renaissance et, d’autre part, les usages que l’on en fait : si Plutarque est considéré comme un auteur moral, voire moralisateur, comme une autorité, on peut constater, chez les auteurs humanistes objets de notre analyse, un usage volontiers hétérodoxe de son œuvre, qu’il s’agisse des Vies ou des Moralia. Ces déviances ou ces écarts par rapport à la doxa nous ont semblé favorisés, voire permis, par les micro-récits que l’on puise chez le Chéronéen et que l’on transplante en sol étranger. Aussi notre travail entreprend-il, à travers la réception de Plutarque à la Renaissance, de « réviser » la catégorie rhétorique de l’exemplum. Les micro-récits plutarquiens nous paraissent se constituer en structures de pensée qui font bouger les normes instituées et autorisent l’émergence d’anomalies. / This study arose from the observation of a gap between, on the one hand, the ethos Plutarch is endowed with in the Renaissance and, on the other hand, the uses that are made of it: if Plutarch is considered to be a moral, even a moralizing, author, and an authority, we can see in the authors of the Renaissance who are the subjects of our analysis a readily unorthodox use of his work, as regards the Parallel Lives or the Moralia. These deviations or discrepancies from the doxa seem to us to have been favored, even allowed, by the short narratives stories that one draws on the Chaeronean works and then transplant in a foreign soil. Therefore our work undertakes, through the reception of Plutarch in the Renaissance, to “revise” the rhetorical category of the exemplum. Plutarchian short narrative stories seem to us to gather structures of thought that change the established standards and allow the emergence of anomalies.

A presente pesquisa aborda o problema das descontinuidades históricas que deram lugar às mudanças epistêmicas ao longo dos séculos. As rupturas cosmológicas são apresentadas como os elementos de erosão epistêmica referidos por Foucault, e apontadas como o gatilho de desencadeamento do conceito de anomalia. Das anomalias celestiais aos indivíduos anormais, questões éticas são evocadas como inquietações que antecedem o que Foucault descreve como a morte do homem. / This research deals with the problem of historical discontinuities that have given rise to epistemic changes through the centuries. Cosmological ruptures are presented as the \"elements of epistemic erosion\" mentioned by Foucault, and identified as unleashing triggers of the anomaly concept. From heavenly anomalies to abnormal individuals, ethical issues are raised as concerns that preced what Foucault describes as the death of man.

Este trabalho tem por objetivo descrever a Anomalia Magnética do Atlântico Sul (SAMA) utilizando os modelos para o período histórico (1600–1890) e também os modelos para o último século (DGRF e IGRF). Como a SAMA apresenta características de baixa intensidade do campo total e coincide com a região de intenso fluxo de partículas cósmicas, muitos problemas com objetos que orbitam a Terra (por exemplo, satélites) são detectados nessa região. São descritos os efeitos provocados pela SAMA em fenômenos espaciais. Através da análise dos modelos na forma de mapas, foram extraídos os dados de mínima intensidade para o centro da SAMA e a posição destes pontos, possibilitando conhecer a trajetória e as taxas de deriva. Os modelos foram testados na interface manto-núcleo através da componente vertical, para encontrar correlação com anomalias em superfície. Os resultados mostraram deriva para Oeste constante e variações em latitude. Foi observado que as intensidades acompanham a diminuição de todo o campo, embora a SAMA apresente caráter predominantemente não-dipolar, evidenciada pela razão entre o campo não-dipolar e o campo total no Atlântico Sul. A comparação de intensidades da SAMA com as medidas de intensidade realizadas pelos observatórios mostrou que a influência da SAMA aparece na forma de sobreposição ou amplificação de fenômenos com menor comprimento de onda, como os impulsos de variação secular (jerks geomagnéticos). A continuação para baixo dos modelos se mostrou satisfatória quando comparada com o método de inversão estocástica. A comparação da SAMA com outras anomalias em superfície mostrou independência na trajetória, porém, quando comparadas com os lóbulos principais na interface manto-núcleo, indicam que estas anomalias possam estar interligadas. Os lóbulos do núcleo foram interpretados com base nos mecanismos de geração, sugerindo que a SAMA possa ser originada através de movimentos combinados entre duas colunas de convecção e regiões de fluxo reverso no núcleo externo. / The object of this dissertation is to describe the South Atlantic Magnetic Anomaly (SAMA) using geomagnetic models for the historical period from 1600 to 1890 and also the IGRF and DGRF models for the past century. Since the SAMA presents low intensities of the total geomagnetic field that correspond to a region of intense cosmic ray particle flux, many problems with objects that orbit along this region (eg. Satelites) have been detected. The SAMA effects on space vehicles are described. The field models led to the definition of the SAMA center as the locus of minimum total field intensity and how the anomaly drifted and varied in intensity for the past four centuries. The vertical component at the Core Mantle Boundary (CMB) was used to find correlations with anomalies at the surface the Earth surface. Results have shown a rather constant westward drift and also latitude variations. It was observed that intensities follow the general decrease of the total field although the SAMA displays a predominantly non-dipolar character that is evident in the non-dipolar/total field ratios for the South Atlantic. The comparison of geomagnetic measurements by nearby Southamerican Observatories show that the SAMA influence appears as a superposition or amplification of lower wavelength phenomena such as the secular variation impulses (jerks). A downward continuation of the models was found satisfactory when compared to the stochastic inversion method. The comparison of the SAMA with other surface anomalies showed a rather independent behavior however, a comparison with the main radial component lobes at the CMB showed that all these anomalies may be interconnected. The nucleus lobes have been interpreted under the light of field generation processes, suggesting that the SAMA may originate from the combined motion of two convection columns and reverse flux regions in the outer core.

Where the role of software-intensive systems has shifted from the traditional one of fulfilling isolated computational tasks, larger collaborative societies with interaction as primary resource, is gradually taking its place. This can be observed in anything from logistics to rescue operations and resource management, numerous services with key-roles in the modern infrastructure. In the light of this new collaborative order, it is imperative that the tools (compilers, debuggers, profilers) and methods (requirements, design, implementation, testing) that supported traditional software engineering values also adjust and extend towards those nurtured by the online instrumentation of software intensive systems. That is, to adjust and to help to avoid situations where limitations in technology and methodology would prevent us from ascertaining the well-being and security of systems that assists our very lives. Coupled with most perspectives on software development and maintenance is one well established member of, and complement to, the development process. Debugging; or the art of discovering, localising, and correcting undesirable behaviours in software-intensive systems, the need for which tend to far outlive development in itself. Debugging is currently performed based on a premise of the developer operating from a god-like perspective. A perspective that implies access and knowledge regarding source code, along with minute control over execution properties. However, the quality as well as accessibility of such information steadily decline with time as requirements, implementation, hardware components and their associated developers, all alike fall behind their continuously evolving surroundings. In this thesis, it is argued that the current practice of software debugging is insufficient, and as precursory action, introduce a technical platform suitable for experimenting with future methods regarding online debugging, maintenance and analysis. An initial implementation of this platform will then be used for experimenting with a simple method that is targeting online observation of software behaviour.

The purpose of this research was to provide a better understanding of meteotsunamis over the eastern Gulf of Mexico along the west coast of Florida and to develop a process for forecasting those events. Meteotsunami waves develop from resonant effects of strong pressure perturbations greater than 1 hPa, moving in excess of 10 m s-1, over water areas up to around 100 m in depth. Meteotsunami events over 0.3 m in height, as measured by three primary NOAA coastal tide gauges at Cedar Key, Clearwater Beach, and Naples, from 2007-2015, impact the Florida Gulf coastline several times per year and are most prevalent south of Cedar Key. Cases that met the indicated thresholds were further examined. A majority of the cases were associated with bands of active convection that brought pressure changes and wind changes. The cases derived from this research provide a baseline for formulating a forecast methodology. The prediction of meteotsunamis is challenging over the marine environment where sub-hourly pressure and wind observations are generally not obtainable. Two forecast methodologies were derived for longer term periods up to several days using numerical model surface pressure data and a refined methodology for forecasts up to several hours in advance of the impacts using a combination of high resolution weather prediction models to provide a robust environment of atmospheric pressure, wind, and pressure fields for prediction of meteotsunamis over shallow shelf waters and available observations. This research illuminates, for National Weather Service forecasters, meteotsunami development and potential hazards related to this phenomenon that can be transmitted to the public within specialized products.

Modern automobiles have been proven vulnerable to hacking by security researchers. By exploiting vulnerabilities in the car's external interfaces, attackers can access a car's controller area network (CAN) bus and cause malicious effects. We seek to detect these attacks on the bus as a last line of defence against automotive cyber attacks. The CAN bus standard defines a low-level message structure, upon which manufacturers layer their own proprietary command protocols; attacks must similarly be tailored for their target. This variability makes intrusion detection methods difficult to apply to the automotive CAN bus. Nevertheless, the bus traffic is generated by machines; thus we hypothesize that it can be characterized with machine learning, and that attacks produce anomalous traffic. Our goals are to show that anomaly detection trained without understanding of the message contents can detect attacks, and to create a framework for understanding how the characteristics of a novel attack can be used to predict its detectability. We developed a model that describes attacks based on their effect on bus traffic, informed by a review of published material on car hacking in combination with analysis of CAN traffic from a 2012 Subaru Impreza. The model specifies three high-level categories of effects: attacks that insert foreign packets, attacks that affect packet timing, and attacks that only modify data within packets. Foreign packet attacks are trivially detectable. For timing-based anomalies, we developed features suitable for one-class classification methods. For packet stream data word anomalies, we adapted recurrent neural networks and multivariate Markov model methods to sequence anomaly detection and compared their performance. We conducted experiments to evaluate our detection methods with special attention to the trade-off between precision and recall, given that a practical system requires a very low false alarm rate. The methods were evaluated by synthesizing anomalies within each attack category, parameterized to adjust their covertness. We generalize from the results to enable prediction of detection rates for new attacks using these methods.

Accurate prediction and control pervades domains such as engineering, physics, chemistry, and biology. Often, it is discovered that the systems under consideration cannot be well represented by linear, periodic nor random data. It has been shown that these systems exhibit deterministic chaos behavior. Deterministic chaos describes systems which are governed by deterministic rules but whose data appear to be random or quasi-periodic distributions. Deterministically chaotic systems characteristically exhibit sensitive dependence upon initial conditions manifested through rapid divergence of states initially close to one another. Due to this characterization, it has been deemed impossible to accurately predict future states of these systems for longer time scales. Fortunately, the deterministic nature of these systems allows for accurate short term predictions, given the dynamics of the system are well understood. This fact has been exploited in the research community and has resulted in various algorithms for short term predictions. Detection of normality in deterministically chaotic systems is critical in understanding the system sufficiently to able to predict future states. Due to the sensitivity to initial conditions, the detection of normal operational states for a deterministically chaotic system can be challenging. The addition of small perturbations to the system, which may result in bifurcation of the normal states, further complicates the problem. The detection of anomalies and prediction of future states of the chaotic system allows for greater understanding of these systems. The goal of this research is to produce methodologies for determining states of normality for deterministically chaotic systems, detection of anomalous behavior, and the more accurate prediction of future states of the system. Additionally, the ability to detect subtle system state changes is discussed. The dissertation addresses these goals by proposing new representational techniques and novel prediction methodologies. The value and efficiency of these methods are explored in various case studies. Presented is an overview of chaotic systems with examples taken from the real world. A representation schema for rapid understanding of the various states of deterministically chaotic systems is presented. This schema is then used to detect anomalies and system state changes. Additionally, a novel prediction methodology which utilizes Lyapunov exponents to facilitate longer term prediction accuracy is presented and compared with other nonlinear prediction methodologies. These novel methodologies are then demonstrated on applications such as wind energy, cyber security and classification of social networks.

The ability to diagnose deviations and predict faults effectively is an important task in various industrial domains for minimizing costs and productivity loss and also conserving environmental resources. However, the majority of the efforts for diagnostics are still carried out by human experts in a time-consuming and expensive manner. Automated data-driven solutions are needed for continuous monitoring of complex systems over time. On the other hand, domain expertise plays a significant role in developing, evaluating, and improving diagnostics and monitoring functions. Therefore, automatically derived solutions must be able to interact with domain experts by taking advantage of available a priori knowledge and by incorporating their feedback into the learning process. This thesis and appended papers tackle the problem of generating a real-world self-monitoring system for continuous monitoring of machines and operations by developing algorithms that can learn data streams and their relations over time and detect anomalies using joint-human machine learning. Throughout this thesis, we have described a number of different approaches, each designed for the needs of a self-monitoring system, and have composed these methods into a coherent framework. More specifically, we presented a two-layer meta-framework, in which the first layer was concerned with learning appropriate data representations and detectinganomalies in an unsupervised fashion, and the second layer aimed at interactively exploiting available expert knowledge in a joint human-machine learning fashion. Furthermore, district heating has been the focus of this thesis as the application domain with the goal of automatically detecting faults and anomalies by comparing heat demands among different groups of customers. We applied and enriched different methods on this domain, which then contributed to the development and improvement of the meta-framework. The contributions that result from the studies included in this work can be summarized into four categories: (1) exploring different data representations that are suitable for the self-monitoring task based on data characteristics and domain knowledge, (2) discovering patterns and groups in data that describe normal behavior of the monitored system/systems, (3) implementing methods to successfully discriminate anomalies from the normal behavior, and (4) incorporating domain knowledge and expert feedback into self-monitoring.

District heating delivers more than 70% of the energy used for heating and domestichot water in Swedish buildings. To stay competitive, district heating needs toreduce its losses and increase capabilities to utilise low grade heat. Finding faultysubstations is one way to allow reductions in supply temperatures in district heatingnetworks, which in turn can help reduce the losses. In this work three suggestedsymptoms of faults: abnormal quantization, drifting and anomalous values, are investigatedwith the help of hourly meter data of: heat load, volume flow, supplyand return temperatures from district heating substations. To identify abnormalquantization, a method is proposed based on Shannon’s entropy, where lower entropysuggests higher risk of abnormal quantization. The majority of the substationsidentified as having abnormal quantization with the proposed method has a meterresolution lower than the majority of the substations in the investigated districtheating network. This lower resolution is likely responsible for identifying thesesubstation, suggesting the method is limited by the meter resolution of the availabledata. To improve result from the method higher resolution and sampling frequencyis likely needed.For identifying drift and anomalous values two methods are proposed, one for eachsymptom. Both methods utilize a software for predicting hourly heat load, volumeflow, supply and return temperatures in individual district heating substations.The method suggested for identifying drift uses the mean value of each predictedand measured quantity during the investigated period. The mean of the prediction iscompared to the mean of the measured values and a large difference would suggestrisk of drift. However this method has not been evaluated due to difficulties infinding a suitable validation method.The proposed method for detecting anomalous values is based on finding anomalousresiduals when comparing the prediction from the prediction software to themeasured values. To find the anomalous residuals the method uses an anomalydetection algorithm called IsolationForest. The method produces rankable lists inwhich substations with risk of anomalies are ranked higher in the lists. Four differentlists where evaluated by an experts. For the two best preforming lists approximatelyhalf of the top 15 substations where classified to contain anomalies by the expertgroup. The proposed method for detecting anomalous values shows promising resultespecially considering how easily the method could be added to a district heatingnetwork. Future work will focus on reducing the number of false positives. Suggestionsfor lowering the false positive rate include, alternations or checks on theprediction models used.

Dimensionality reduction techniques such as t-SNE and UMAP are useful both for overview of high-dimensional datasets and as part of a machine learning pipeline. These techniques create a non-parametric model of the manifold by fitting a density kernel about each data point using the distances to its k-nearest neighbors. In dense regions, this approach works well, but in sparse regions, it tends to draw unrelated points into the nearest cluster. Our work focuses on a homotopy method which imposes graph-based regularization over the manifold parameters to update the embedding. As the homotopy parameter increases, so does the cost of modeling different scales between adjacent neighborhoods. This gradually imposes a more uniform scale over the manifold, resulting in a more faithful embedding which preserves structure in dense areas while pushing sparse anomalous points outward.