FID4SA-Repository
Not a member yet
21181 research outputs found
Sort by
Le banc de l'amitié par Béatrice Mabilon-BonfilsLe laboratoire du bonheur
No full textchronique radio et article blo
Hirao Cross-Coupling Reaction as an Efficient Tool to Build Non-natural C2-Phosphonylated Sugars
No full textInternational audienceAbstract A range of C2-phosphonylated sugars have been accessed through a palladium-catalyzed Hirao cross-coupling on 2-iodoglycals using trialkylphosphites as phosphorylating reagents. The developed conditions led to the creation of an unnatural C–P bond on sugars and proved to be compatible with diversely protected glycals (acetyl-, benzyl-, PMB-protected) as well as with unprotected substrates. Several monosaccharides and one disaccharide have been synthesized by applying this methodology. Deprotection conditions are also described
Deep-Learning Technology for Book Conservation Assessment in Libraries and Archives
No full textInternational audienceOne of the primary goals of Libraries and Archives is the conservation of their collections in order to pass them on to future generations. The large number of books kept in storage makes this task particularly challenging. Artificial intelligence offers the possibility of processing large volumes of data in a short period of time, but this technology has not been used for book conservation. A team of artificial intelligence scientists and one conservator is now developing a tool to automatically assess the conservation state of a binding at any given time
Des pratiques aux apprentissages mathématiques, en passant par la formation.: Circulation des savoirs issus des recherches en Didactique des Mathématiques
No full textThe synthesis of my research work presented here aims at linking a posteriori several axes of reflection around the analysis of teaching practices in mathematics, axes pursued thanks to different collaborations with other researchers since my thesis. While remaining within the framework of the Didactic and Ergonomic Double Approach for the analysis of teaching practices, and still giving a central place to the mathematical activities of the students, our tools to describe and characterise what teachers do in mathematics class have been enriched, through contact with other approaches, and in particular through the consideration of questions relating to the evaluation of students' mathematical learning. On the other hand, the passage from local case studies, repeated many times over on different mathematical contents, to large-scale studies, in terms of the number of teachers involved or the long duration of the study, has enabled us, not only to consolidate our analytical tools, by highlighting what, in our opinion, can play a role in the mathematical learning of pupils, but also to confirm some of our results on a smaller scale concerning the variety and coherence of practices. Finally, these results feed our reflections on teacher training, and equip us to analyse the training systems and practices of teacher educators, as well as their effects on the development of teaching practices in mathematics. This synthesis concludes with possible research to be carried out in the wake of these results.La synthèse de mes travaux de recherche présentée ici vise à relier a posteriori plusieurs axes de réflexion autour de l’analyse de pratiques enseignantes en mathématiques, axes poursuivis grâce à différentes collaborations avec d’autres chercheur.e.s depuis ma thèse. Tout en restant dans le cadre de la Double Approche Didactique et Ergonomique pour l’analyse des pratiques enseignantes, et en donnant toujours une place centrale aux activités mathématiques des élèves, nos outils pour décrire et caractériser ce que font les enseignant.e.s en classe de mathématiques se sont enrichis, au contact d’autres approches, et en particulier à travers la prise en compte de questions relatives à l’évaluation des apprentissages mathématiques des élèves. D’autre part, le passage d’études de cas locales, maintes fois répétées sur différents contenus mathématiques, à des études à grande échelle, en ce qui concerne les effectifs d’enseignant.e.s impliqué.e.s ou le temps long de l’étude, nous a permis non seulement de consolider nos outils d’analyse, en mettant en lumière ce qui, selon nous, peut jouer un rôle dans les apprentissages mathématiques des élèves, mais aussi de confirmer certains de nos résultats à plus petite échelle sur la variété et la cohérence des pratiques. Enfin, ces résultats nourrissent nos réflexions sur la formation des enseignant.e.s, et nous outillent pour analyser les dispositifs de formation et les pratiques des formateurs et formatrices, ainsi que leurs effets sur le développement des pratiques enseignantes en mathématiques. Des pistes de recherches à mener dans le prolongement de celles-ci concluent cette synthèse
The Environmental Impact of Internet Regulation
No full textWe address the need to regulate Internet infrastructure usage to take into account environmental externalities. We model the interactions between a monopoly ISP and different types of content providers in settings where the former chooses the network size and the latter influences congestion on the network. We first show that current net neutrality regulation does not provide agents the right incentives to cope with the environmental externality issue. Then, we study several alternatives, including laissez-faire, price-based regulation, and norm-based regulation. We derive conditions under which these alternatives fare better than net neutrality. In particular, the two types of regulations are useful tools to accommodate consumer interest and environmental concerns
On Predictive Explanation of Data Anomalies
No full textNumerous algorithms have been proposed for detecting anomalies (outliers, novelties) in an unsupervised manner. Unfortunately, it is not trivial, in general, to understand why a given sample (record) is labelled as an anomaly and thus diagnose its root causes. We propose the following reduced-dimensionality, surrogate model approach to explain detector decisions: approximate the detection model with another one that employs only a small subset of features. Subsequently, samples can be visualized in this low-dimensionality space for human understanding. To this end, we develop PROTEUS, an AutoML pipeline to produce the surrogate model, specifically designed for feature selection on imbalanced datasets. The PROTEUS surrogate model can not only explain the training data, but also the out-of-sample (unseen) data. In other words, PROTEUS produces predictive explanations by approximating the decision surface of an unsupervised detector. PROTEUS is designed to return an accurate estimate of out-of-sample predictive performance to serve as a metric of the quality of the approximation. Computational experiments confirm the efficacy of PROTEUS to produce predictive explanations for different families of detectors and to reliably estimate their predictive performance in unseen data. Unlike several ad-hoc feature importance methods, PROTEUS is robust to high-dimensional data
Bagnes coloniaux de Guyane et de Nouvelle-Calédonie (XIXe-XXe siècles)
No full textInternational audienc