374 research outputs found
Qui sont les Dii mauri ?
No full textThe author reviews the various but rare interpretations of the divine collectivity referred to by the name of Dii Mauri. From a re-examination of the literary or epigraphic mentions of the 50 specifically African local or regional divinities, and of the 20 dedications to the Dii Mauri known to this day, he believes he can propose the relationship, even the identity, between the local gods and the Dii Mauri. The comparison between the authors of the dedications shows that the worship of the local gods principally concerned « civilians » (82,5 %), whereas the Dii Mauri were invoked by governors, imperial procurators, soldiers (81,25 %). Moreover the Dii Mauri are invoked as often in Numidia and in Africa as in Caesarian Mauretania (they are unknown in Tingitana). Therefore the term of "maurus" is not linked to the Roman administrative carving, it applies to what is rebellious to Latin culture, to what is specifically native and unassimilable. Dii Mauri and African gods are the same divinities, only the dedicators change.The author reviews the various but rare interpretations of the divine collectivity referred to by the name of Dii Mauri. From a re-examination of the literary or epigraphic mentions of the 50 specifically African local or regional divinities, and of the 20 dedications to the Dii Mauri known to this day, he believes he can propose the relationship, even the identity, between the local gods and the Dii Mauri. The comparison between the authors of the dedications shows that the worship of the local gods principally concerned « civilians » (82,5 %), whereas the Dii Mauri were invoked by governors, imperial procurators, soldiers (81,25 %). Moreover the Dii Mauri are invoked as often in Numidia and in Africa as in Caesarian Mauretania (they are unknown in Tingitana). Therefore the term of "maurus" is not linked to the Roman administrative carving, it applies to what is rebellious to Latin culture, to what is specifically native and unassimilable. Dii Mauri and African gods are the same divinities, only the dedicators change.Camps Gabriel. Qui sont les Dii mauri ?. In: Antiquités africaines, 26,1990. pp. 131-153
Modeling Threats to AI-ML Systems Using STRIDE
No full textThe application of emerging technologies, such as Artificial Intelligence (AI), entails risks that need to be addressed to ensure secure and trustworthy socio-technical infrastructures. Machine Learning (ML), the most developed subfield of AI, allows for improved decision-making processes. However, ML models exhibit specific vulnerabilities that conventional IT systems are not subject to. As systems incorporating ML components become increasingly pervasive, the need to provide security practitioners with threat modeling tailored to the specific AI-ML pipeline is of paramount importance. Currently, there exist no well-established approach accounting for the entire ML life-cycle in the identification and analysis of threats targeting ML techniques. In this paper, we propose an asset-centered methodology—STRIDE-AI—for assessing the security of AI-ML-based systems. We discuss how to apply the FMEA process to identify how assets generated and used at different stages of the ML life-cycle may fail. By adapting Microsoft’s STRIDE approach to the AI-ML domain, we map potential ML failure modes to threats and security properties these threats may endanger. The proposed methodology can assist ML practitioners in choosing the most effective security controls to protect ML assets. We illustrate STRIDE-AI with the help of a real-world use case selected from the TOREADOR H2020 project
STRIDE-AI: An Approach to Identifying Vulnerabilities of Machine Learning Assets
No full textWe propose a security methodology for Machine Learning (ML) pipelines, supporting the definition of key security properties of ML assets, the identification of threats to them as well as the selection, test and verification of security controls. Our proposal is based on STRIDE, a widely used approach to threat modeling originally developed by Microsoft. We adapt STRIDE to the Artificial Intelligence domain by taking a security property-driven approach that also provides guidance in selecting the security controls needed to alleviate the identified threats. Our proposal is illustrated via an industrial case study
Hardening behavioral classifiers against polymorphic malware: An ensemble approach based on minority report
No full textIn recent years, malware attacks have become more and more sophisticated, reflecting a radical change in malware behavior. Attackers aim to create malware that, at each execution, generates a different number of independent and cooperating threads. Randomization of malware's division of labor among threads poses significant challenges to traditional detection approaches. In this paper, we demonstrate that attacks based on random division of labor among multiple threads can dramatically degrade the detection performance of five benchmark ML models, in some cases dropping their accuracy to 50% with only a few threads. Then, we propose and evaluate a novel detection technique based on polymorphic-aware training and ensemble learning with ad-hoc voting scheme (favoring minority report). Results of experimentation carried out on real malware system call logs and assigned to threads via a Bayesian splitting accounting for inter-call dependency indicate that our ensemble has high detection capabilities (99.7% best case), and improves the baseline accuracy of a single model in detecting single-thread malware
Tradução comentada do conto Lizards in Jamshyd's Courtyard, de William Faulkner
No full textDissertação (mestrado) - Universidade Federal de Santa Catarina, Centro de Comunicação e Expressão. Programa de Pós-graduação em Estudos da TraduçãoEste trabalho de dissertação é fruto de estudos de teorias da tradução e teve como princípio norteador a aquisição de conhecimentos sobre aspectos relacionados com a produção da obra original, para só então definir a posição do tradutor. Somente após a contextualização da obra original e análise das características do autor concretizou-se a tradução do conto Lizards in Jamshyd's Courtyard de William Faulkner. Para manter a força do conto original não houve simplesmente a preocupação em conseguir encontrar equivalentes ou traduzir palavra por palavra, mas sim, em adentrar no jogo de significantes, de maneira a tornar a tradução o mais próximo possível do original, respeitando a heterogeneidade das situações lingüísticas e culturais existentes entre a língua inglesa do original e a língua portuguesa no Brasil, para a qual o conto foi traduzido. Muitos obstáculos foram encontrados ao longo desse processo, e a estes, foram apresentadas soluções. Tanto as hipóteses levantadas para a solução dos problemas, quanto as decisões tomadas descritas nesta pesquisa estão ancoradas nos princípios teóricos de Lawrence Venutti, Georges Mounin, John C. Catford e Antoine Berman. This essay has its origins in studies about translation theories and in the knowledge acquisition about the aspects related with the production of the original work. Just after those studies, was established the position as translator. And only after the contextualization of the original work and the analysis of the author characteristics it was started the translation process of the tale Lizards in Jamshyd's Courtyard written by William Faulkner, this tale is part of the novel Hamlet written by the same author. To maintain the strength of the original tale there was not just a concern about getting equivalents or translating word by word , but was to be very close to the characteristics of the original tale; considering what is heterogeneous in the linguistic and cultural situations between the English language in which the original tale was written, and the Portuguese language from Brazil where the tale has been translated. The hypothesis, the possible solutions to the problems found, and the decisions taken in this research are based on: Lawrence Venutti, Georges Mounin, John C. Cattford and Antoine Berman's theories
Tihei Mauri Ora: A Māori response to health disparities
No full textMāori bear a disproportionate burden of health problems which, in concert with other factors (e.g. poor housing, low socio-economic status and low education attainment), contribute to and maintain low health status. It is noted that there have been multiple attempts to reduce health inequities – however, such attempts have been largely unsuccessful. Barriers to success include government reticence, restrictions on Māori participation in determining health directions/solutions, current contract paradigms and a reluctance to engage in meaningful partnerships with Māori. Those barriers occur within a cultural framework which defines (and therefore prioritises) the health of an individual over the needs of the collective.
The hypothesis of this research is that Māori health disparities are best addressed via the development and delivery of Māori health models by services which are oriented to kaupapa Māori principles. Utilising a case study approach, this thesis looks at the outcomes generated when a kaupapa Māori service applies key Māori principles to health service delivery. The case study, in tandem with focus group interviews identifies the key elements necessary to developing services which are responsive to the needs of Māori.
This study identified the importance of promoting change (and ultimately improve Māori health status) that encompasses the formation of a framework which considers collective benefit over individualism, encompasses Māori values, acknowledges and accepts Māoricentric clinical interventions. In addition, the thesis asserts that Māori health status will improve once Māori are active participants rather than recipients of health services
DATA PARTITIONING AND COMPENSATION TECHNIQUES FOR SECURE TRAINING OF MACHINE LEARNING MODELS
No full textAdvances in Machine Learning (ML), coupled with increased availability of huge amounts of data collected from diverse sources and improvements in computing power, have led to a widespread adoption of ML-based solutions in critical application scenarios. However, ML models intrinsically introduce new security vulnerabilities within the systems into which they are integrated, thereby expanding their attack surface. The security of ML-based systems hinges on the robustness of the ML model employed. By interfering with any of the phases of the learning process, an adversary can manipulate data and prevent the model from learning the correct correlations or mislead it into taking potentially harmful actions. Adversarial ML is a recent research field that addresses two specific research topics. One of them concerns the identification of security issues related to the use of ML models, and the other concerns the design of defense mechanisms to prevent or mitigate the detrimental effects of attacks.
In this dissertation, we investigate how to improve the resilience of ML models against training-time attacks under black-box knowledge assumption on both the attacker and the defender. The main contribution of this work is a novel defense mechanism which combines ensemble models (an approach traditionally used only for increasing the generalization capabilities of the model) and security risk analysis. Specifically, the results from the risk analysis in the input data space are used to guide the partitioning of the training data via an unsupervised technique. Then, we employ an ensemble of models, each trained on a different partition, and combine their output based on a majority voting mechanism to obtain the final prediction. Experiments are carried out on a publicly available dataset to assess the effectiveness of the proposed method. This novel defence technique is complemented by two other contributions, which respectively support using a Distributed Ledger to make training data tampering less convenient for attackers, and using a quantitative index to compute ML models’ performance degradation before and after the deployment of the defense. Taken together, this set of techniques provides a framework to improve the robustness of the ML lifecycle
Dragan Jakovljević, Erkenntnisgestalten und Handlungsanweisungen. Abhandlungen zur Erkenntnislehre und praktischen Philosophie, hrsg. von H.R. Sepp, Verlag Traugott Bautz, Nordhausen 2016 («Libri Nigri», Bd. 57). Un volume di pp. 201
No full textRenouncing authentic thought and arguing on the basis of abstract categories and sterile juxtapositions of concepts is a risk to which every thinker is subject. The philosophical debate itself often takes place within conceptual paradigms which, although universally accepted, sometimes do not correspond to the actual situation they are supposed to describe and need further examination. Dragan Jakovljević's book, which takes the form of a collection of seven different essays, most of which have already been published previously, constantly strives to unmask the false myths on which philosophical discussion in various spheres has been guiltily fossilized in recent years. The author - a professor of ethics and theory of knowledge at the University of Podgorica - offers a sharp and balanced reflection on some of the hottest topics in various disciplines, in an analysis that ranges from epistemology to ethics, from social sciences to philosophy of religion. Thus, the epistemological issues of the validity of knowledge and the method of scientific research are addressed, as well as problems that emerge from the public debate, such as the correct conception of tolerance in a pluralistic society or the role that religion can and must play in this context. The argumentation is mostly based on a comparison with the tradition of critical rationalism, in particular the thinking of K.R. Popper and H. Albert. Albert, with whom the author completed his doctoral studies at the University of Mannheim
Securing Machine Learning Models: Notions and Open Issues
No full textMachine Learning (ML) models are taking the place of conventional algorithms in a wide range of application domains. However, once ML models have been deployed in the field, they can be attacked in ways that are very different from the ones of conventional systems. This chapter reviews some of the techniques that attackers use to compromise ML-based systems at two core phases of the learning process: the training and the inference stages. It provides an overview that, taking into account the current variety and scope of threats and attacks to ML models, will help the security analyst in charge of alleviating them. The chapter introduces some preliminary concepts, including the one of ML lifecycle. It then presents the setting of Adversarial Machine Learning from the point of view context of computer security, and discusses the notions of threats, vulnerabilities, and attacks. The chapter also details common alleviation measures against training-time attacks
Attività di impiego e di testing di armi anti-satellite e diritto internazionale
No full textAnti-satellite weapons (ASAT) are one of the major challenges to international security in the «Fourth Domain», namely outer space. This was proved by the test made by Russia in November 2021 and the reaction of the international community. The present testing of said arms against its own satellites, as well as the future and probable operating use of said arms against satellites of other States, raises the problem of their compatibility with international law. In particular, it is being debated in various international fora the legality of ASATs, whose use can generate space debris further congesting terrestrial orbits, thus interfering (and jeopardizing) the activities of other States.
The article concentrates on the existing rules of space law (jus ad bellum), (jus in bello) and environmental law in order to define the limits imposed on such activities. Then the author reviews the very recent phenomenon of unilateral declarations, by some States, of renunciation to test direct-ascent, kinetic ASAT, in order to assess if such renunciation may constitute the embryo of evolving customary law
- …
