L'analyse de séries temporelles multivariées, créées par des capteurs présents sur le moteur d'avion durant un vol ou un essai, représente un nouveau challenge pour les experts métier en aéronautique. Chaque série temporelle peut être décomposée de manière univariée en une succession de phases transitoires, très connues par les experts, et de phases stabilisées qui sont moins explorées bien qu'elles apportent beaucoup d'informations sur le fonctionnement d'un moteur. Notre projet a pour but de convertir ces séries temporelles en une succession de labels, désignant des phases transitoires et stabilisées dans un contexte bivarié. Cette transformation des données donne lieu à plusieurs perspectives : repérer dans un contexte univarié ou bivarié les patterns similaires durant un vol, trouver des tronçons de courbes similaires à une courbe donnée, identifier les phases atypiques, détecter ses séquences de labels fréquents et rares durant un vol, trouver le vol le plus représentatif et déterminer les vols «volages». Ce manuscrit propose une méthodologie pour automatiquement identifier les phases transitoires et stabilisées, classer les phases transitoires, labelliser des séries temporelles et les analyser. Tous les algorithmes sont appliqués à des données de vols et les résultats sont validés par les experts. / Analysing multivariate time series created by sensors during a flight or a bench test represents a new challenge for aircraft engineers. Each time series can be decomposed univariately into a series of stabilised phases, well known by the expert, and transient phases that are merely explored but very informative when the engine is running. Our project aims at converting these time series into a succession of labels, designing transient and stabilised phases in a bivariate context. This transformation of the data will allow several perspectives: tracking similar behaviours or bivariate patterns seen during a flight, finding similar curves from a given curve, identifying the atypical curves, detecting frequent or rare sequences of labels during a flight, discovering hidden multivariate structures, modelling a representative flight, and spotting unusual flights. This manuscript proposes : methodology to automatically identify transient and stabilized phases, cluster all engine transient phases, label multivariate time series and analyse them. All algorithms are applied on real flight measurements with a validation of the results from expert knowledge.

This study was designed to determine the prevalence of substance use and risky sexual behaviours among university students in Ethiopia and examine whether personality traits predict substance use and risky sexual behaviours. Using a cross-sectional quantitative design, data were obtained from 2620 undergraduate students selected through stratified multistage sampling. Personality traits were measured through John, Donahue, and Kentle‘s (1991) Big Five Inventories [BFI-44]. Substance use and risky sexual behaviours were assessed using risky behaviour scales adapted from Miller et al. (2004) and Zuckerman and Kuhlman (2000). Statistical analysis was performed using Statistical Package for Social Sciences (SPSS) Version 20 and Analysis of Moment Structure (AMOS) Version 22 software. The findings revealed that 72.18% of the respondents reported that they have used at least one kind of substance in their lifetime and 50.27% of the participants had consumed at least one kind of substance in the past 30 days before the survey. The current prevalence of substance use was 46.3%, 16.1%, 9.9%, 6.8%, and 5.0% for drinking alcohol, chewing khat, smoking shisha, smoking cigarettes and marijuana use respectively. About 40.2% of respondents had sexual intercourse at least once in their lifetime and 25.6% had sex in the past three months before the study. Among sexually active respondents, 35% began sexual intercourse at the age of 17 years old or younger; 64.3% had multiple sexual partners; 53.1% had substance-induced sex, 62.03% had unsafe sex, and 45.3% had casual sex at least once. Males were found to be more at risk of substance use and risky sexual behaviours. Analysis of the structural relation revealed that conscientiousness and agreeableness traits were significant negative predictors and extraversion was a significant positive predictor of substance use and risky sexual behaviours. Neuroticism had a significant direct effect only on substance use. Substance use mediated the effect of personality traits on risky sexual behaviours. There was gender moderation or variation on the effect of personality traits on substance use and risky sexual behaviours. Therefore, the findings may imply that health risk behaviours were highly prevalent among students, which requires special prevention and intervention. Personality traits can be used to identify the vulnerable individuals and design programs aimed at developing behaviours underlying the protective traits. / Psychology / D. Litt. et Phil. (Psychology)

Le point de départ de cette thèse a été l'expression de «culture du risque» telle qu'elle a été employée par l'État suite aux événements liés à la tempête Xynthia en 2010. À travers celle-ci, ce dernier rend compte d'un manque de connaissances et de comportements inadaptés, caractérisant les populations vis-à-vis des aléas côtiers. Parallèlement, il affiche sa volonté de favoriser leur sensibilisation vis-à-vis de ces problématiques. Cependant, nous avons constaté qu'un dispositif législatif de communication et de sensibilisation existe depuis les années 1980, mais que son bilan paraît toujours mitigé. De plus, deux visions principales paraissent s'affronter : d'un côté, l'État qui cherche à limiter l'implantation des enjeux dans des zones dites « à risque », de l'autre les populations locales qui demandent à protéger leurs enjeux. Ainsi, la confrontation aux mêmes aléas ne donne pas lieu à des interprétations identiques : différentes cultures du risque existent. Afin de comprendre ces différences, nous tenterons de démontrer l'hypothèse stipulant que les différents discours observés auprès des individus en matière de risques côtiers sont déterminés par leurs rapports au milieu. Il sera donc question de caractériser ces rapports et d'illustrer en quoi ceux-ci sont susceptibles d'affecter leurs discours et leurs comportements à propos des aléas côtiers. Parallèlement, nous tenterons de définir les composantes d'une culture du risque et d'en évaluer le niveau auprès des résidents oléronais. Pour cela, nous nous appuierons sur l'analyse statistique de résultats issus d'une enquête par questionnaire, mais aussi sur l'analyse de cartes mentales. / The starting point of this thesis was the expression "risk culture" as used by the French government following the events related to the storm Xynthia in 2010. Through this, it reports a lack of knowledge and inappropriate behaviours among the populations when faced with coastal hazards. At the same time, it demonstrates its desire to raise public awareness of these issues. However, we have noted that a legislative communication and awareness-­raising framework has been in place since the 1980s, but its results are still limited. In addition, two main visions seem to be confronting each other: on the one hand, the State, which seeks to limit the establishment of stakes in so-called "at-risk" areas, and on the other hand, the local inhabitants, who as to protect their stakes. Thus, confronting the same hazards does not lead to identical interpretations: different cultures of risk exist. In order to understand these differences, we will attempt to demonstrate the hypothesis that the different discourses observed among populations regarding coastal risks are determined by their relationship to the environment. The aim is to characterise these relationships and illustrate how they are likely to affect their discourses and behaviours about coastal hazards. At the same time, we will attempt to define the components of a risk culture and assess its level among Oleronian residents. To do this, we will rely on the statistical analysis of results from a questionnaire survey, but also on mental maps.

No description available.

Lors du développement d’une tumeur au sein d’un tissu, les cellules cancéreuses se retrouvent entourées par les cellules saines. Les interactions entre ces deux types cellulaires, transformé et normal, jouent un rôle important dans le devenir de la tumeur, mais restent à ce jour mal comprises. L’objectif de cette thèse a été de mettre en place des systèmes in vitro qui permettent d’étudier les interactions entre une population de cellules normales et une population de cellules transformées.Nous avons tiré profit d’une lignée de cellules épithéliales sensibles à la lumière, élaborée par Olivier Destaing (IAB, Grenoble). Lorsqu’elles sont exposées à la lumière bleue, ces cellules suractivent la protéine Src, connue pour être surexprimée dans de nombreux cancers. Sinon, elles gardent un phénotype normal. L’utilisation de ces cellules, appelées « OptoSrc », combinée à un dispositif optique, permet de créer des tissus mosaïques dans lesquels le motif des cellules mutées est déterminé par le motif d’illumination bleue. Notre système présente plusieurs avantages : le contrôle dans le temps et dans l’espace du motif de cellules transformées, mais aussi l’activation graduelle et réversible de l’oncoprotéine.Nous avons montré qu’en illuminant dans le bleu un îlot circulaire de cellules au sein d’une monocouche OptoSrc, les cellules activées s’extrudent collectivement, donnant naissance à un agrégat tri-dimensionnel cohésif surplombant la monocouche. Nous pouvons contrôler la taille et le temps d’apparition de ce sphéroïde en ajustant respectivement l’aire éclairée et la fréquence d’illumination. De plus, ce phénomène d’extrusion collective est réversible lorsque le stimulus de lumière bleue s’arrête. Finalement, nous avons montré que la formation de cet agrégat s’accompagne d’une diminution des E-cadhérines à la membrane, et de l’apparition de la vimentine, pour les cellules éclairées. Nos résultats suggèrent qu'un groupe de cellules surexprimant la protéine Src, au sein d’une monocouche de cellules normales, subit une transition epithéliale- mesenchymateuse partielle. / During the development of a tumour in a tissue, the cancer cells are surrounded by healthy cells. The interactions between these two cell types, transformed and normal, play an important role in the tumour stability, but remain to this day poorly understood. The aim of this thesis was to establish in vitro assays to study the interactions between populations of normal and transformed cells.We benefited from a light-sensitive cell line, constructed by Olivier Destaing (IAB, Grenoble). When they are exposed to blue light, these cells overactivate the protein Src, which is known to be overexpressed in many cancers. Otherwise, they keep a normal phenotype. Using these cells, called “OptoSrc”, in combination with an optical setup, we are able to create mosaic tissues in which the pattern of mutated cells is determined by the blue illumination pattern. Our system has several advantages: a selective control in time and space of the group of transformed cells, and a gradual and reversible activation of the oncoprotein.We have shown that when we illuminate a circular islet of cells from a monolayer of OptoSrc cells, the activated cells were collectively extruded, resulting in a cohesive three-dimensional aggregate on top of the monolayer. We can control the size and appearance time of this spheroid by tuning, respectively, the area and frequency of illumination. Besides, this collective extrusion is reversible when the blue light stimulation is stopped. Finally, we have shown that the formation of this three-dimensional aggregate coincides with the loss of E-cadherin at the membrane, as well as the apparition of vimentin, for the illuminated OptoSrc cells. Our results suggest that a group of cells overexpressing the protein Src, in a monolayer of normal cells, undergoes a partial epithelial-to-mesenchymal transition.

L’évaluation des émotions positives chez l’animal est encore délicate. Pourtant, de nombreuses évidences démontrent l’intérêt de favoriser l’expression de ces émotions, notamment dans le but d’améliorer le bien-être des individus. Cette recherche vise à valider et identifier des indicateurs d’émotions positives chez le cheval. Dans un premier temps, une étude expérimentale basée sur l’utilisation d’accéléromètres, tout comme l’observation d’épisodes d’anticipation alimentaire quotidiens en centre équestre, ont permis d’identifier le déclenchement de l’anticipation et de mieux caractériser les comportements exprimés dans ce cadre. Nos résultats dévoilent que les comportements d’anticipation reflètent davantage un état de frustration. Ils ne constituent donc pas un indicateur fiable d’émotions positives. En revanche, une augmentation d’intensité de ces comportements semble indiquer un état de mal-être. Puis, en nous appuyant sur l’utilisation d’un casque permettant la mesure de l’activité électroencéphalographique chez des chevaux éveillés et libres de leurs mouvements, nous avons pu (i) confirmer l’importance de l’hémisphère gauche dans le traitement des émotions positives et de l’hémisphère droit dans le traitement attentionnel, (ii) mettre en évidence le rôle majeur des ondes thêta lors d’un traitement cognitif attentionnel, (iii) souligner l’interaction des processus émotionnels et attentionnels chez le cheval. Par ailleurs, si le test classique d’attention visuelle (VAT) constitue un bon outil de mesure des capacités attentionnelles d’un cheval au travail, ce même test réalisé sans humain apparait être un outil novateur de mesure de l’état de bien-être. Enfin, l’étude de la production de sons non-vocaux, basée sur une approche comparative combinant l’observation de chevaux de centre équestre et des chevaux vivant en condition semi-naturelle (favorable au bien-être), a révélé que l’ébrouement (hors contrainte respiratoire liée à une maladie ou à un équipement mal ajusté) serait un marqueur fiable d’émotions positives d’intensité modérée chez le cheval, puisqu’il exprimerait que ce dernier perçoit de façon positive son environnement. L’ensemble des recherches menées dans ce travail ouvrent vers des perspectives d’applications intéressantes dans le milieu équin, mais aussi d’un point de vue fondamental et méthodologique. / Indicators of positive emotions in animals are still scarce. However, many studies showed that favour these emotions is a promising way to improve individual welfare state. This study aims to validate and identify some indicators of positive emotions in horses. First, based both on an experimental study using accelerometers and the observation of riding school horses during their regular feeding time, we characterized anticipatory behaviours in this specie and showed that they were triggered by an environmental cue. Moreover, anticipatory behaviours were associated with frustration and thus were not reliable indicators of positive emotions. However, an increase of their intensity may reflect a poor welfare state. Then, the use of an EEG helmet designed for awake horses free from their movements, allowed us to (i) confirm the left-hemisphere bias for positive emotions processes and the right hemisphere bias for attentional processes, (ii) highlight the implication of theta waves in cognitive attentional processes, (iii) underline interactions between attentional and emotional processes in horses. Moreover, we showed that the measures of attentional characteristics thanks to a visual attention test (VAT) was useful to define attentional capacities at work. The same test conducted without a human was rather useful to evaluate the horses’ welfare state. Finally, a comparative approach based on the observation of horses living either in restricted or in naturalistic conditions revealed that snort (out of breathing difficulties due to illness or poorly fitting equipment) could be a reliable indicator of positive emotions of low intensity as it would be the expression of the horses’ positive appraisal of a situation. This works opens up promising fundamental and applied perspectives and applications.

Overweight, a common health condition in adolescence, has been linked with difficulties in emotional processing. This study investigates associations between emotional processing, conceptualised through the model of Emotional Openness (EO), problematic eating behaviours, including Eating in the Absence of Hunger and disinhibited eating, and overweight in adolescents. Several self-report instruments were completed by 160 youngsters (mean age: 14.36 ± 0.61 years) from the community, including 39 overweight and obese adolescents (24.5%). In girls, bootstrap analyses supported a mediating effect of restrained eating on the relation between three EO dimensions and body mass index percentile, in particular the communication of emotions, the cognitive-conceptual representation of emotions, and the perception of bodily indicator of emotions. No mediating effect was found in boys. These results have important implications for psychological weight management interventions, as they underline the relevance of work on emotional processing in order to reduce problematic eating behaviours.

This essay examined how bilingual peace education can promote peaceful behaviours and social integration among otherwise segregated Arab- Israelis and Jewish- Israelis. By using The School for Peace at Neve Shalom/ Wahat al-Salam as a case study, accompanied with Jürgen Habermas theory of communicative action and New institutionalism with a conflict critical approach, we observed the behavioural mechanisms of institutions and social interactions. The empiric data consisted of three interviews conducted by author Nava Sonnenschein from the book The Power of Dialogue between Israelis and Palestinians. The interviewed students from The School for Peace had experienced both conventional education and bilingual peace education in adulthood. We found that informal values and narratives in educational institution influence the behaviours of its students. We also found that interactions between Arab- Israelisand Jewish- Israelis in the context of bilingual peace education promoted recognition of commonalities. While conventional education generally formed hostile behaviours and segregating incitements, the bilingual peace education gave opportunities for interactions and acknowledgement of both ethnic groups; consequently, leading to self-reflection, mutual understanding, peaceful behaviours and social integration.

Diploma thesis deals with the current methods of implementation of primary prevention of primary schools in the Czech Republic. The aim was to determine the degree of preparedness of primary schools for the implementation of a world-wide research-proven program of positive support for students behavior in the school environment (PBIS) in order to prevent risky behavior of students in schools. The first part of the thesis is devoted to theoretical knowledge about students with risky behavior, preventive and interventional procedures of primary prevention applied in schools and a thorough description of the PBIS method. We used a qualitative method of semi-structured interviews for the research survey. The evaluation took place on the basis of interviews with pedagogical staff at four primary schools. The analysis was performed using open coding. Questions were asked about the use of preventive measures by primary schools and the answers provided an insight into which elements used correspond to the PBIS method. The results of the research show similarities in the use of preventive elements in the field of physical environment of the class and in the positive formulation of the rules of conduct. On the contrary, visualizations of routine procedures and school-wide expectations in school premises are...

Vector Symbolic Architectures (VSA) combine a hypervector space and a set of operations on these vectors. Hypervectors provide powerful and noise-robust representations and VSAs are associated with promising theoretical properties for approaching high-level cognitive tasks. However, a major drawback of VSAs is the lack of opportunities to learn them from training data. Their power is merely an effect of good (and elaborate) design rather than learning. We exploit high-level knowledge about the structure of reactive robot problems to learn a VSA based on training data. We demonstrate preliminary results on a simple navigation task. Given a successful demonstration of a navigation run by pairs of sensor input and actuator output, the system learns a single hypervector that encodes this reactive behaviour. When executing (and combining) such VSA-based behaviours, the advantages of hypervectors (i.e. the representational power and robustness to noise) are preserved. Moreover, a particular beauty of this approach is that it can learn encodings for behaviours that have exactly the same form (a hypervector) no matter how complex the sensor input or the behaviours are.