25353 research outputs found
Sort by
Regio-and Chemoselective Double Allylic Substitution of Alkenyl vic-Diols
No full textInternational audienceDouble allylic substitution is an attractive approach to build molecular complexity from simple starting materials by creating two new bonds in one-pot. However, this type of reaction has been doomed by chemoselectivity and regioselectivity issues. In this manuscript, we describe a new approach to introduce a-la-carte two new C-C, C-N, C-O or C-S bonds in a chemo-and regioselective fashion. The reaction relies on sequential dual catalysis with Lewis acid and palladium. The scope is remarkably broad, and the reaction can be diastereoselective using secondary alcohols as the first nucleophile.</div
Detection and suppression of epileptiform seizures via model-free control and derivatives in a noisy environment
No full textInternational audienceRecent advances in control theory yield closedloop neurostimulations for suppressing epileptiform seizures. These advances are illustrated by computer experiments which are easy to implement and to tune. The feedback synthesis is provided by an intelligent proportional-derivative (iPD) regulator associated to model-free control. This approach has already been successfully exploited in many concrete situations in engineering, since no precise computational modeling is needed. iPDs permit tracking a large variety of signals including high-amplitude epileptic activity. Those unpredictable pathological brain oscillations should be detected in order to avoid continuous stimulation, which might induce detrimental side effects. This is achieved by introducing a data mining method based on the maxima of the recorded signals. The real-time derivative estimation in a particularly noisy epileptiform environment is made possible due to a newly developed algebraic differentiator. The virtual patient is the Wendling model, i.e., a set of ordinary differential equations adapted from the Jansen-Rit neural mass model in order to generate epileptiform activity via appropriate values of excitation-and inhibition-related parameters. Several simulations, which lead to a large variety of possible scenarios, are discussed. They show the robustness of our control synthesis with respect to different virtual patients and external disturbances
Flooding as a sub-critical instability in open channels
No full textInternational audienceIn flood events caused by a gradual increase in the flow rate of a watercourse, the rise in water level is often abrupt, while the fall in level is delayed. We show that such behavior can be demonstrated by considering stationary flows at high Reynolds number in a prismatic open channel: several geometries of the channel cross-section lead to a subcritical instability that results in a discontinuous rise in the level when the flow rate exceeds a critical value F, and in a fall, also discontinuous, when the flow rate returns below a value F lower than F. This hysteretic behavior originates from the interplay between gravity which drives the flow downstream, and turbulent friction with the channel wall. The potential existence of several solutions arising from this bifurcation requires careful consideration in flood simulations
Modèles génératifs pour le traitement des données du type électrocardiogramme : théorie et application.
No full textThis thesis contributes to the vast domain of Generative models, with a particular interest in applying such models to electrocardiogram (ECG) data for inference and uncertainty quantification.In a first part, we develop two novel methods for reducing bias in Importance Sampling and Sequential Monte Carlo (SMC) methods, which are two important tools of Bayesian inference.The issuing algorithms can both be viewed as a wrapper around current existing algorithms providing effortless bias reduction. We also provide new non-assymptotic convergence bounds for using such algorithms for parameter learning in Hidden Markov Models (HMM).In a second part, we focus on using SMC for solving Bayesian linear inverse problems with generative models serving as informative priors.Finally, we apply this method on several ECG based inverse problems, namely missing lead completion and out-of-distribution detection.Cette thèse apporte des contributions au vaste domaine des modèles génératifs, avec un intérêt particulier pour l'application de tels modèles aux données d'électrocardiogramme (ECG) dans le cadre de l'inférence et de la quantification de l'incertitude.Dans une première partie, nous développons deux méthodes novatrices pour réduire le biais dans les méthodes d'échantillonnage d'importance et de Monte Carlo séquentiel (SMC), qui sont deux outils importants de l'inférence bayésienne. Les algorithmes résultants peuvent être considérés tous deux comme des "enveloppes" autour d'algorithmes existants actuels, offrant une réduction de biais sans grande augmentation du temps de calcul.Nous présentons également de nouvelles bornes de convergence non asymptotiques pour l'utilisation de ces algorithmes dans l'apprentissage de paramètres dans les modèles de Markov cachés (HMM).Dans une deuxième partie, nous nous concentrons sur l'utilisation du SMC pour résoudre des problèmes inverses linéaires bayésiens, avec des modèles génératifs servant de priors informatifs. Cette approche est particulièrement intéressante pour améliorer la résolution des problèmes inverses rencontrés dans divers domaines scientifiques.Enfin, nous appliquons cette méthodologie à plusieurs problèmes inverses basés sur l'ECG, notamment la complétion de pistes manquantes et la détection hors distribution.Les résultats de ces applications démontrent l'efficacité et la polyvalence des modèles génératifs proposés pour relever des défis concrets dans le contexte de l'analyse des données ECG
Property-Based Testing by Elaborating Proof Outlines
No full textInternational audienceProperty-based testing (PBT) is a technique for validating code against an executable specification by automatically generating test-data. We present a proof-theoretical reconstruction of this style of testing for relational specifications and employ the Foundational Proof Certificate framework to describe test generators. We do this by encoding certain kinds of ``proof outlines'' as proof certificates that can describe various common generation strategies in the PBT literature, ranging from random to exhaustive, including their combination. We also address the shrinking of counterexamples as a first step toward their explanation. Once generation is accomplished, the testing phase is a standard logic programming search. After illustrating our techniques on simple, first-order (algebraic) data structures, we lift it to data structures containing bindings by using the -tree syntax approach to encode bindings. The Prolog programming language can perform both generating and checking of tests using this approach to syntax. We then further extend PBT to specifications in a fragment of linear logic. Under consideration in Theory and Practice of Logic Programming (TPLP)
How could 50 °C be reached in Paris: Analyzing the CMIP6 ensemble to design storylines for adaptation
No full textInternational audienceReaching a surface temperature of 50 °C in a heavily populated region, like Paris, would have devastating effects. Although such a high value seems far from the present-day record of 42.6 °C, its occurrence cannot be dismissed by the end of the 21st century, due to the continuous increase of global mean temperature. In this paper, we address two questions that were asked by the City of Paris to a group of scientists: When does this event start to be likely? What are the prevailing meteorological conditions? We base our study on the CMIP6 simulation ensemble. Many of the CMIP6 yield biases in temperature. Rather than using methods of bias correction, which are not necessarily adapted to high extremes, we propose a pragmatic approach of model selection in order to seek such high temperature events that are deemed realistic. We analyze the meteorological conditions leading to first occurrences of such hot events and their common atmospheric patterns. This paper describes a simple data mining approach (on a large ensemble of climate model simulations) which could be adapted to other regions of the world, in order to help decision makers anticipating and adapting to such devastating meteorological events
TECHNOLOGY FOR CHANGE: Inclusive innovation and inclusion practices in companies
No full textInternational audienceLa Chaire Technology for Change, soutenue par un programme de mécénat entre Accenture et l'Institut Polytechnique de Paris, a pour but d'examiner et de renforcer les liens entre les technologies et le développement durable, incluant ses dimensions sociales, économiques et environnementales. L'objectif des travaux de la Chaire est de suggérer des voies permettant aux technologies de jouer un rôle dans la résolution de différentes problématiques sociales et environnementales, dont celle de l'exclusion. C'est dans ce but que la Chaire Technology for Change soutient les actions de l'Observatoire de l'innovation inclusive depuis sa création.Le TechLab d'APF France handicap promeut l'innovation inclusive dans tous les processus de conception de produits et services. Il accompagne les entreprises dans leurs démarches de co-conception avec des personnes en situation de handicap, des aidants et des professionnels du réseau APF France handicap. Depuis sa création, le TechLab inscrit sa pratique dans une logique de création et de partage de connaissances sur l'innovation inclusive et sur les défis qu'elle pose aux organisations.Accenture Research dessine les tendances et crée des points de vue s'appuyant sur des analyses de données avancées. Associant la puissance de méthodes de recherches innovantes et sa connaissance poussée des industries, notre équipe de 300 chercheurs et analystes est implantée dans 20 pays et publie chaque année des rapports, articles et points de vue. Adossée à des données exclusives et à des partenariats, notre analyse des tendances suscite la réflexion, guide nos innovations et nous permet de transformer des théories et des idées novatrices en solutions concrètes pour les organisations.La Fondation Accenture agit depuis plus de 25 ans en France en faveur de la formation, de l'emploi et de l'inclusion, en s'appuyant sur l'expertise technologique et de conseil des collaborateurs d'Accenture
α-Amido Trifluoromethyl Xanthates: A New Class of RAFT/MADIX Agents
No full textInternational audienceXanthates have long been described as poor RAFT/MADIX agents for styrene polymerization. Through the determination of chain transfer constants to xanthates, this work demonstrated beneficial capto-dative substituent effects for the leaving group of a new series of α-amido trifluoromethyl xanthates, with the best effect observed with trifluoroacetyl group. The previously observed Z-group activation with a O-trifluoroethyl group compared to the O-ethyl counterpart was quantitatively established with Cex = 2.7 (3–4 fold increase) using the SEC peak resolution method. This study further confirmed the advantageous incorporation of trifluoromethyl substituents to activate xanthates in radical chain transfer processes and contributed to identify the most reactive xanthate reported to date for RAFT/MADIX polymerization of styrene
Ressources humaines et éthique du capitalisme ? [Eric Godelier]
No full textXerfi Canal spoke to Eric Godelier, professor at the Ecole Polytechnique, about the absence of ethical capitalism.Interview conducted by Jean-Philippe Denis.Xerfi Canal a reçu Eric Godelier, professeur à l’Ecole Polytechnique, pour parler de l'absence de capitalisme éthique.Une interview menée par Jean-Philippe Denis