HAL-Ecole des Ponts ParisTech
Not a member yet
    42743 research outputs found

    Policy with guaranteed risk-adjusted performance for multistage stochastic linear problems

    No full text
    International audienceRisk-averse multi-stage problems and their applications are gaining interest in various fields of applications. Under convexity assumptions, the resolution of these problems can be done with trajectory following dynamic programming algorithms like Stochastic Dual Dynamic Programming (SDDP) to access a deterministic lower bound, and dual SDDP for deterministic upper bounds. In this paper, we leverage the dual SDDP algorithm to compute a policy with guaranteed risk-adjusted performance for multistage stochastic linear problems

    L'amibiase au Liban : prévalence, susceptibilité génétique, et corrélation d'Entamoeba histolytica avec la qualité de l'eau

    No full text
    Entamoeba is a genus of parasitic protozoa that includes Entamoeba histolytica (E. histolytica), known for causing amebiasis in humans. In South Lebanon, high rates of amebiasis among infants and children imply the presence of E. histolytica in water sources. Contaminated tap water, used for hygiene, poses a risk, particularly since this parasite can persist in water for months, influenced by environmental conditions. Our study was undertaken in Nabatieh and Hasbayya districts to evaluate domestic water quality regarding parasitic contamination and to correlate water quality variables with parasitic presence, while improving detection methods.In this research, 90 drinking water samples and 15 surface water samples were analyzed for various parameters, including physical, chemical, and microbiological qualities. The contamination by Entamoeba spp. was assessed using microscopy and a new spectrophotometric method. Notably, significant contamination levels of phosphates, coliforms, and Entamoeba spp. were found in Nabatieh's domestic water. Additionally, contamination in the Hasbani river and domestic water in Hasbayya was confirmed, particularly with total and fecal coliforms and Entamoeba spp. cysts, posing health risks to children in the region. A weak correlation between cysts and total coliforms suggested possible groundwater contamination by soil.Further, an epidemiological study with 180 children revealed a 25% prevalence of amebiasis, connecting consumption of contaminated water and agricultural practices near homes to increased risk. Despite these findings, community awareness regarding E. histolytica remains low.Potentially, high prevalence rates may be overestimated due to analytical inaccuracies or genetic predisposition among local children. Traditional microscopy is insufficient for species differentiation and may produce false positives; therefore, molecular diagnostics are pivotal for accurate disease understanding in the region.Collected stool samples from diagnosed children were re-analyzed with species-specific PCR, confirming the molecular prevalence of amebiasis. A case-control genetic association study also encompassed 180 participants, assessing the TLR2 rs3804099 polymorphism related to amebiasis susceptibility. Results indicated a strong association between this genetic variant and increased disease risk, underlining the influence of genotype-phenotype correlations on infectious disease susceptibility in endemic areas. This study emphasizes the critical role of water quality in health and highlights the necessity for advanced diagnostic methods in accurately assessing the prevalence of E. histolytica.Entamoeba est un genre de protozoaires parasites, dont l'espèce la plus connue, Entamoeba histolytica, cause l'amibiase chez les humains. Au sud du Liban, une forte prévalence de l'amibiase chez les jeunes enfants indique la présence de ce pathogène d'origine hydrique dans les sources d'eau locales. L'eau du robinet contaminée, utilisée pour l'hygiène corporelle, peut être ingérée par les enfants. Étant donné que ce parasite peut survivre dans l'eau pendant plusieurs mois selon la température, la qualité de l'eau est cruciale pour la santé publique. Les objectifs de cette étude étaient d'évaluer la qualité de l'eau domestique, d'examiner la corrélation entre contamination parasitaire et variables de qualité de l'eau, et d'améliorer les méthodes de détection.Nos études ont été menées dans les districts de Nabatiyeh et Hasbayya, en tenant compte des différences environnementales entre villes et villages. Au total, 90 échantillons d'eau potable et 15 d'eau de surface ont été analysés pour leurs paramètres physiques, chimiques et bactériologiques. La contamination par Entamoeba spp. a été examinée par microscopie, tandis qu'une analyse spectrophotométrique a été effectuée pour identifier des pics communs parmi les échantillons positifs. Pour la première fois, la contamination de l'eau domestique à Nabatiyeh par des phosphates, des coliformes totaux et fécaux, ainsi que par des kystes d'Entamoeba spp. a été confirmée. À Hasbayya, des contaminants microbiologiques similaires ont été trouvés dans l'eau de la rivière Hasbani et dans l'eau domestique, indiquant un risque pour la santé des enfants.Une faible corrélation a été observée entre les kystes et les coliformes totaux, suggérant une contamination potentielle des réseaux d'eau souterrains par le sol. Parallèlement, une méthode simple et économique a été développée pour mesurer la concentration de kystes d'Entamoeba spp., ce qui pourrait faciliter la recherche.Pour mieux comprendre la prévalence et les risques environnementaux associés à l'amibiase à Nabatiyeh, une étude épidémiologique impliquant 180 enfants a été réalisée. Les résultats ont révélé une prévalence de 25% d'amibiase, avec des facteurs de risque liés à la consommation d'eau contaminée et aux activités agricoles à proximité des logements. La sensibilisation insuffisante de la population concernant E. histolytica a également été mise en lumière.Il convient de noter que la forte prévalence de l'amibiase pourrait également être liée à des erreurs analytiques et à une prédisposition génétique. La microscopie optique, largement utilisée dans les centres médicaux, peut conduire à des faux positifs, tandis que le diagnostic moléculaire serait plus spécifique. Les variations génétiques des gènes de réponse immunitaire, notamment les récepteurs TLR, peuvent jouer un rôle dans cette dynamique.Dans le cadre de l'étude, 40 échantillons de selles ont été collectés des enfants diagnostiqués avec une infection à E. histolytica et analysés par PCR. Une étude de cas-témoins, comprenant 180 participants, a exploré le polymorphisme rs3804099 du gène TLR2. Les résultats indiquent une forte prévalence d'amibiase et une association entre le polymorphisme et un risque accru chez les enfants Libanais, soulignant l'importance de méthodes de diagnostic plus précises et la nécessité de comprendre les implications des facteurs génétiques sur la santé des enfants dans les régions touchées

    Theoretical Analysis of Branch Prediction in the Sliding Window Algorithm

    No full text
    We revisit the classical sliding window pattern matching algorithm from a computer architecture point of view and conduct a theoretical study to answer the question: what is the impact of adding a simple branch prediction mechanism to the model of computation? Assuming that the pattern is fixed and the text is random, we give precise estimations of the number of mispredictions produced by two variants of this algorithm using local predictors. This case study introduces key tools based on automata and Markov chain theories, which are promising for the analysis of more efficient algorithms and more advanced branch prediction schemes

    Stabilisation et contrôle des systèmes dynamiques : des méthodes classiques à l’apprentissage par renforcement.

    No full text
    This thesis explores the application of reinforcement learning (RL) and control theory in evolution systems such as traffic stabilization, pest control, and fluid dynamics. It begins with a theoretical study on the exponential stability of the Saint-Venant system with viscosity under appropriate boundary control, by constructing a quadratic Lyapunov function. The need for a new Lyapunov function arises as the non-viscous case does not extend to the viscous system, improving the understanding of real-world flow behavior where viscosity is ubiquitous.The thesis then focuses on applying RL to biological pest control, training a neural network to stabilize mosquito populations by optimizing the release rate of sterile insects. A mathematical analysis follows, yielding an explicit control law derived from the RLtrained model—an innovative approach for extracting mathematical controls for complex systems with limited observations.Finally, RL is applied to traffic stabilization, where controllers for autonomous vehicles (AVs) are designed and trained in simulation to smooth out stop-and-go waves. Even a small proportion of AVs can significantly reduce fuel consumption and CO 2 emissions for all road users. These RL controllers are then deployed in the real world to control 100 vehicles in live, busy highway traffic as part of the CIRCLES Consortium’s MegaVanderTest, the largest AV-based traffic-smoothing experiment to date.Overall, this work highlights the versatility of RL and control theory in solving real-world stabilization problems. It bridges the gap between theoretical stability analysis and large-scale deployment in critical applications such as traffic and biological control.Cette thèse explore l’application de l’apprentissage par renforcement (AR) et de la théorie du contrôle à des systèmes évolutifs comme la stabilisation du trafic, la lutte biologique contre les nuisibles et la dynamique des fluides. Elle commence par une étude théorique de la stabilité exponentielle du système de Saint-Venant avec viscosité sous un contrôle aux frontières adapté, en construisant une fonction de Lyapunov quadratique. Le cas non visqueux ne s’étendant pas au système visqueux, une nouvelle fonction de Lyapunov est nécessaire, améliorant ainsi la compréhension des écoulements réels où la viscosité est omniprésente.La thèse s’intéresse ensuite à l’application de l’AR à la lutte biologique, où un réseau de neurones est entraîné pour stabiliser les populations de moustiques en optimisant le lâcher d’insectes stériles. Une analyse mathématique en découle, aboutissant à une loi de contrôle explicite dérivée du modèle entraîné par AR—une approche innovante pour extraire des contrôles mathématiques de systèmes complexes avec des observations limitées.Enfin, l’AR est appliqué à la stabilisation du trafic, avec des contrôleurs pour véhicules autonomes (VA) conçus et entraînés en simulation pour atténuer les ondes d’accordéon. Même une faible proportion de VA réduit considérablement la consommation de carburant et les émissions de CO 2 pour l’ensemble des usagers de la route. Ces contrôleurs sont ensuite déployés en conditions réelles pour piloter 100 véhicules sur une autoroute en heure de pointe, dans le cadre du MegaVanderTest du Consortium CIRCLES, la plus grande expérimentation de stabilisation du trafic par VA à ce jour

    Time Consistency for Multistage Stochastic Optimization Problems under Constraints in Expectation

    No full text
    International audienceWe consider sequences-indexed by time (discrete stages)-of families of multistage stochastic optimization problems. At each time, the optimization problems in a family are parameterized by some quantities (initial states, constraint levels.. .). In this framework, we introduce an adapted notion of time consistent optimal solutions, that is, solutions that remain optimal after truncation of the past and that are optimal for any values of the parameters. We link this time consistency notion with the concept of state variable in Markov Decision Processes for a class of multistage stochastic optimization problems incorporating state constraints at the final time, either formulated in expectation or in probability. For such problems, when the primitive noise random process is stagewise independent and takes a finite number of values, we show that time consistent solutions can be obtained by considering a finite dimensional state variable. We illustrate our results on a simple dam management problem

    Optimization and Beyond

    No full text
    International audienceThis paper will be concerned with hard choices—that is, choice situations where an agent cannot make a rationally justified choice. Specifically, this paper asks: if an agent cannot optimize in a given situation, are they facing a hard choice? A pair of claims are defended in light of this question. First, situations where an agent cannot optimize because of incompleteness of the binary preference or value relation constitute a hard choice. Second, situations where agents cannot optimize because the binary preference or value relation violates acyclicity do not constitute a hard choice

    Mechanical behaviour and microstructure of methane hydrate-bearing sandy sediment observed at various spatial scales

    No full text
    International audienceMethane hydrates (MHs) are considered an alternative energy resource but also a potential source of geo-hazards and climate change. The physical/mechanical properties of gas hydrate-bearing sandy sediments are strongly dependent on the distribution of hydrates within the pore space. The purpose of this study is to investigate morphologies and pore-habits of MHs formed in sandy sediments by means of experiments that probe a wide range of scales, from the pore scale – using Synchrotron X-Ray Computed Tomography (SXRCT) and optical microscopy – to the core scale, through mechanical property measurements. The same synthetic sands are used, in which MHs are generated successively under excess gas and excess water conditions. At the macroscopic (core) scale, MH pore habits are inferred by comparing the measured sonic wave velocities to velocities calculated from rock physics models and further assessed via triaxial compression tests. Furthermore, Magnetic Resonance Imaging is used to investigate the kinetics of MH formation and distribution along the core height. The pore habits and MH morphologies are directly visualized at the pore (grain) scale by SXRCT and, with still better spatial and temporal resolution, by transmission optical microscopy, revealing some more complex morphologies than in the hydrate pore habits commonly admitted

    On the micro- and meso-structure and durability of 3D printed concrete elements

    No full text
    International audience3D printed concrete (3DPC) creates opportunities, including a reduction in construction waste and time and increased design freedom. However, because of the differences in the construction technique compared to traditional concrete casting, the structures also perform differently; namely, the micro- and meso-structure and durability are shown to be different. For the 3DP technology to find its way to the market, one needs to be aware of these differences and needs to know how to quantify the above-mentioned properties, as differences in the testing methodologies impose themselves when characterizing printed instead of cast concrete. In this paper, we elaborate on the test methods to investigate the micro- and meso-structure and the durability of 3DPC. We start with a discussion on the micro- and meso-structure of the 3D printed concrete and how it is different from conventional mold-cast concrete. An in-depth discussion of the test methods to assess the durability of 3D printed concrete is outlined. Reported findings related to the two aforementioned properties are discussed. In addition, we report on the technologies proposed to improve the durability performance of 3DPC, and we highlight the remaining challenges and opportunities related to 3DPC

    RILEM Contribution to Earthen Building

    No full text
    International audienceEarth as building materials and earthen building products are important for the conservation and rehabilitation of historic and vernacular construction, but also to build new eco-efficient buildings. Earth is not standardized as for other more common construction materials. For earthen products to be applied, depending on the building technologies used and the architectural and engineering design, their performance must be known. Therefore, there is a need to define standardized test procedures, so that performance results can be validated and compared. Within the International Union of Laboratories and Experts in Construction Materials, Systems and Structures (RILEM) an effort has been made, so that earthen materials and products can be addressed, and test methods can be discussed, optimised and defined. This article presents the evolution within RILEM of earth as building material, earthen building products and elements, details difficulties, but also the achievements reached so far. It intends to spread the work now being done, namely within RILEM Technical Committees BEC, MAE and PEM, so that a larger number of professionals can contribute and profit from results achieved

    5,758

    full texts

    42,743

    metadata records
    Updated in last 30 days.
    HAL-Ecole des Ponts ParisTech
    Access Repository Dashboard
    Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇