HAL-Ecole des Ponts ParisTech
Not a member yet
42743 research outputs found
Sort by
Contributions on complexity bounds for Deterministic Partially Observed Markov Decision Process
No full textInternational audienceMarkov Decision Processes (Mdps) form a versatile framework used to model a wide range of optimization problems. The Mdp model consists of sets of states, actions, time steps, rewards, and probability transitions. When in a given state and at a given time, the decision maker's action generates a reward and determines the state at the next time step according to the probability transition function. However, Mdps assume that the decision maker knows the state of the controlled dynamical system. Hence, when one needs to optimize controlled dynamical systems under partial observation, one often turns toward the formalism of Partially Observed Markov Decision Processes (Pomdp). Pomdps are often untractable in the general case as Dynamic Programming suffers from the curse of dimensionality. Instead of focusing on the general Pomdps, we present a subclass where transitions and observations mappings are deterministic: Deterministic Partially Observed Markov Decision Processes (Det-Pomdp). That subclass of problems has been studied by (Littman, 1996) and (Bonet, 2009). It was first considered as a limit case of Pomdps by Littman, mainly used to illustrate the complexity of Pomdps when considering as few sources of uncertainties as possible. In this paper, we improve on Littman's complexity bounds. We then introduce and study an even simpler class: Separated Det-Pomdps and give some new complexity bounds for this class. This new class of problems uses a property of the dynamics and observation to push back the curse of dimensionality
Migrations résidentielles vers les espaces ruraux et lieu d’emploi. Un exode urbain à relativiser ?
No full textInternational audienceQuestionnement scientifique : Ces dernières années, outre les médias, certains rapports confirmeraient un retournement dans l’attractivité résidentielle des métropoles (Eurofound and European Commission Joint Research Centre, 2024). De plus en plus d’habitant·es chercheraient à quitter les métropoles pour des espaces ruraux (Milet et al., 2022). Pourtant, la logique de concentration des emplois pour une plus grande efficacité dans un monde globalisé est toujours prégnante parmi les politiques et les chercheurs (Davezies and Estèbe, 2014). Dans cette logique, les équipements et les services se concentrent dans les villes les plus grandes dans un souci théorique de restriction des dépenses publiques. On peut alors s’interroger sur la contradiction urbaine que cela peut créer et les questions de mobilité que cela pose.Les pratiques de mobilité des nouveaux·elles habitant·es des espaces ruraux en provenance d’une métropole sont liées à leurs trajectoires résidentielles. En effet, les socialisations antérieures et les contextes de résidence influencent les mobilités quotidiennes tout au long de la vie (Cailly et al., 2020). L'arrivée de "néo-ruraux·ales" peut ainsi transformer les dynamiques de mobilité sur le territoire, en initiant de nouvelles pratiques (Cailly, 2024). Par ailleurs, Batisse et al., mettent en avant l’influence des migrations résidentielles sur la ségrégation sociale dans les espaces ruraux. Les dynamiques de mobilité des néoruraux·ales et des liens avec leur ancien lieu de travail restent largement à définir (Champion et al., 2009).Notre recherche s’intéresse ainsi aux pratiques de mobilités des personnes actives qui ont quitté les métropoles pour s’installer dans les espaces ruraux et les petites villes en France au regard de leur catégorie socio-professionnelle et de leur situation familiale. Nous cherchons à comprendre qui sont ces nouveaux·elles ruraux·ales, quel rapport ils·elles entretiennent avec leur ancienne métropole de résidence et quelles sont leurs pratiques de mobilité quotidienne dans leur nouvel espace de résidence, analysées au regard du lieu d’emploi et à l’échelle du ménage. Méthodologie et analyse : Les méthodes d’analyses quantitative permettent de mettre en évidence ces phénomènes. Nous utilisons le recensement de la population française de 2023 en fusionnant différentes bases spécialisées : les migrations résidentielles, la mobilité quotidienne et la structure du ménage avec les liens entre les personnes d’un même ménage.Pour différentier les différents types de territoires, nous utilisons les aires d’attraction de l’INSEE comme découpage territorial avec d’un côté les métropoles françaises et l’Île-de-France et de l’autre les espaces ruraux et les petites villes. Évidemment, les aires d’attractions contiennent des espaces à dominante rurale, pour autant nous nous intéressons aux migrations vers des espaces à priori hors de toute attraction métropolitaine, c’est pourquoi nous ne garderons pas dans l’analyse les migrations vers des territoires ruraux qui appartiennent à des aires d’attraction des plus grandes villes.Les nouveaux·elles résident·es sont identifié·es comme les actif·ves qui ont déménagé depuis moins de 5 ans. À partir de modèles de régression logistique et multilinéaire, nous tenterons d’expliquer les choix résidentiels et les spécificités de pratiques de mobilité liées au lieu de travail en fonction des caractéristiques socio-démographiques des ménages.Principaux résultats attendus : Nous souhaitons dans un premier temps analyser ces migrations résidentielles au regard du lieu d’emploi. Un emménagement en espace rural ou dans une petite ville se traduit par des réalités différentes en termes de lien à la métropole d’origine : certain·es actif·ves conservent un emploi dans la métropole lorsque d’autres habitent et travail en dehors de la métropole. Nous soulignons que ces différents types de migrations sont corrélés à la catégorie professionnelle.Dans un deuxième temps, il semble important de s’intéresser aux couples de bi-actif·ves qui ont déménagé. Emménager en espace rural ou dans une petite ville peut être le fruit d’un arbitrage dans les membres du couple, conduisant à des distances au lieu d’emploi déséquilibrées. Nous souhaitons interroger la dimension genrée de ces arbitrages. Ces analyses permettent dans un troisième temps de mieux appréhender les mobilités domicile-travail des nouveaux·elles actif·ves ayant vécu dans une métropole. Nous souhaitons comparer leurs pratiques à celles des ancien·nes résident·es des espaces ruraux. Cela nous permettra de discuter de diverses formes de mobilité entre transition mobilitaire et ancrage dans les espaces ruraux et les petites villes. Ces nouvelles pratiques illustrent-elles un renforcement de l’autonomie de ces espaces ou une dynamique de dépendance aux métropoles
Water penetration in the microstructure of hardwood revealed by NMR relaxometry
No full textInternational audienceThe imbibition of water in the microstructure of cubic hardwood (oak and poplar) samples was followed by NMR, with sample faces either open to air or coated with paint, and along the different wood directions. Dynamic NMR relaxometry allows to clearly distinguish and quantify the amount of water appearing in fibers, in vessels, or as bound water, over time. It appears that the water penetrates first in the form of bound water in the wood sample. Subsequently, fibers are infiltrated at a slower rate, followed by vessels, which exhibit the slowest rate of invasion. For poplar, vessels even start to be invaded only after all the fibers have been filled. Furthermore, the invasion dynamics of the different phases are qualitatively similar when all open faces of the wood sample are coated with paint, preventing any air extraction by these faces. A simple capillary imbibition model fails to fully describe these processes, indicating that the wetting properties vary depending on the presence of bound water in the cell walls, and subsequently, on the presence of water in fibers. Finally, given that bound water penetrates prior to free water, the diffusion coefficient of bound water can be estimated from the data across different directions (L, R, T), which enables the characterization of moisture exchange between construction materials and ambient air under hygroscopic conditions
Rheology of Liquid Foam flowing through Porous Media
No full textInternational audienceLiquid foams are widely used in porous media, like soils. Recent studies have explored foam flow in two-and three-dimensional porous media, but the effects of confinement on foam rheology remain poorly understood. We investigated foam flow through granular packing with stable bubble sizes and reduced compressibility across different liquid fractions. Our analysis shows that confined foam flow cannot be directly extrapolated from bulk foam properties. The apparent yield stress of confined foam increases with degree of confinement, governed by the bubble-to-grain size ratio. This increase can be significant, especially for foams with high liquid content. The increase aligns with a smooth transition between yield stress values of bulk foams and those under strong confinement. This suggests that the geometric effect of bubbles passing through small constrictions is not significant. Instead, liquid redistribution within pore spaces at grain surfaces and contact areas is crucial. This effect is modeled and agrees with our data and literature. Finally, stopping bubble coarsening is shown to be essential to accurately understand foam rheology in porous media
Attributing the occurrence and intensity of extreme events with the flow analogue method
No full textInternational audienceAbstract. Extreme event attribution methodologies have been proposed to estimate the impacts of anthropogenic global warming on observed climatological and meteorological extremes. The classical risk-based approach uses extreme value theory (EVT) to derive changes in the unconditional probabilities of yearly maxima but bears the risk of comparing events with different dynamical mechanisms. The flow analogue method is a conditional attribution method which compares events with similar synoptic-scale dynamics. Here we propose a procedure for estimating both the conditional intensity change and the probability ratio of observed extreme events with this method. We illustrate the procedure on three recent extreme events in Europe and compare the results obtained to the EVT-based approach. We show that the conditional flow analogue method tends to give more significant results for these events, which suggests a stronger climate change signal than the one detected with the unconditional approach
Characterization of Atmospheric Visibility through Extinction Coefficient and the Influence of Lower Threshold on Assessment of Multifractal Parameters
No full textInternational audienceScaling analysis of subjectively defined fields, which are often expressed as a range of values based on the field of application, is not straightforward. To avoid potential biases in statistical analysis, extracting the actual underlying field and understanding the effect of ranged values are important. One such example is atmospheric visibility which is often estimated as a range in meteorological context, as meteorological optical range (MOR). This is estimated from the extinction coefficient σ e , an objective measurement of light attenuation by constituent gases and aerosols in the atmosphere, and expressed as a range depending on the application needs. Accurate estimation of visibility and its variability is required for the safe functioning of various domains such as transport sectors and free optic communication or for understanding regional variations in air quality and climate. Since MOR is a subjective range, here we attempt to characterize it using the objectively measured σ e . In this context, we identify and illustrate the effect of a lower threshold in the data, which is not exclusive to the current problem, and examine its consequences in the multifractal characterization of the field. Here, σ e was extracted from visibility data by a present weather sensor located in the Paris region (France). This was then compared with σ e extracted from MOR measured at Paris Charles de Gaulle airport during the same period. Variability in σ e was investigated under the framework of universal multifractals (UMs), which is widely used to characterize geophysical fields that exhibit extreme variability across scales. With direct data analysis and numerical simulations mimicking the behavior, it was found that the multifractal properties exhibited by σ e are influenced by the upper limit of visibility range in the data. The biases are identified within the theoretical framework of UM, thus expanding the general understanding on the retrieval of the underlying unbiased stochastic field. Significance Statement Data obtained from measurement campaigns are often influenced by instrumental limitations such as upper and lower detection limits. One such example is atmospheric visibility which is usually represented as a range, from objective measurement of light attenuation by constituent particles. Here, we examine the consequences of characterizing visibility using the extinction coefficient, by considering the presence of a lower limiting value, using two datasets in the Paris region—forward scattering sensor and airport visibility measurement. Through data analysis, numerical simulation, and theoretical formulation, the biases are identified and the underlying unbiased stochastic field is retrieved. This enables a more accurate analysis and simulation of visibility, as well as other geophysical fields with similar instrumental or application limitations
Sex and Gender Analysis as a scientific practice: methodological, practical, and ethical challenges
No full textInternational audienceSex and Gender Analysis (SGA) refers to several practices to account for sex and/or gender differences in scientific research, including the use of biological sex and/or gender as a variable and related analysis of the differences across sexes and genders, or at least the disaggregation of data by sex/gender. Motivated by its envisaged and documented benefits both for science and the broader society, efforts to introduce SGA in scientific practice have been quite successful: now many research funders and funding schemes require researchers to include SGA in their research protocols. Yet, the implementation of SGA policies faces many methodological, practical, and ethical challenges. First, to perform sound statistical analysis, the inclusion of biological sex as well as other factors in experiment design must be coupled with appropriate statistical power. Without enough subjects, experiments designed to analyse many different biological variables at once are more likely to reject the null hypothesis. Solving this problem by, for instance, increasing the number of mice in a preclinical study comes with extra methodological issues (e.g. more chances to get false positives), practical issues (e.g. more money needed), and ethical issues (e.g. more mice lives needed). Second, in some cases, it might be difficult to apply SGA to the concrete reality of a lab. For instance, in cell culture research it can be difficult to obtain cells from female and male donors3. Some even question whether single cells from male or female donors can reasonably be considered representative of male or female sex given the very different conditions between a cell in a living sexed organism and a cell in an in vitro culture. Third, from a policy perspective, full implementation of SGA requires major changes in the incentive and reward system. Currently, many agencies require researchers to include SGA in their research proposals, but not many agencies check if researchers actually perform those analyses after obtaining funding. Without further changes, SGA might remain a buzzword in research proposals with no influence on scientific practice. In general, how to evaluate the implementation of SGA policy is still unclear. A recent study showed that many funding agencies lack guidance on the evaluation of such policies. Fourth, while SGA is supposed to deliver better health treatment to discriminated groups such as women, it is worse suited to address health injustice towards other marginalized groups. The widespread use of biological sex as a binary variable and the mix-up of the terms sex and gender in the medical literature clash with the many criticisms of and evidence against the concept of sex as binary. This means that without further changes in biomedical education, SGA might improve the way female patients are treated, but it could not bring much improvement in the lives of, for instance, intersex people.Here, I reconstruct the recent history of SGA and review the main open issues and challenges regarding its implementation as a scientific practice discussed in the literature. These are not necessarily reasons to dismiss SGA, but rather an opportunity for practice-oriented philosophers of science to give their contribution to understanding SGA as a scientific practice and perhaps improve on its implementation
Behavioral and computational signatures of reinforcement learning and confidence biases in gambling disorder
No full textInternational audienceBackground and aimsGambling Disorder (GD) is associated with maladaptive decision-making, possibly driven by biases in learning and confidence judgments. While prior research report abnormal learning rates and heightened overconfidence in GD, the affected cognitive mechanism producing these joint deficits has so far remained unidentified. Our study aims to fill this gap using a recently established reinforcement learning (RL) experimental and computational framework linking learning processes, outcome-valence effects and confidence judgments.MethodsWe pre-registered and tested the hypotheses that GD patients exhibit increased (over)confidence and confirmatory learning bias, and increased outcome valence effects on choice accuracy and confidence judgements in in 18 participants with GD and 19 matched controls.ResultsWhile our findings replicated the main behavioral patterns of choices and confidence judgments, and confirmed their computational foundations, we did not find any group differences between the controls and patients with GD.Discussion and ConclusionsThe current findings speak to the inconsistent findings of abnormalities in confidence and learning in GD. Systematic research is necessary to better understand the influence of possibly mediating factors such as disorder-related idiosyncrasies (e.g. skill- vs chance-based preferences) to further clarify if, when and how confidence and learning are affected in people with GD
Intertwined renewable and digital transitions: a study on South Australia’s hybridized electricity system
No full textInternational audienceThis paper examines the implications of an energy transition characterized by the widespread adoption of renewable energy sources, distributed energy assets, and digitalization, using South Australia as a central case study. By leveraging and adapting the concept of “Electrical Hybridization” (Verdeil and Jaglin, 2023), we show how South Australia’s ambitious energy policies, which integrate economic growth with decarbonization, are transforming its electricity system. This transformation requires both the reinforcement of the electricity network (reticulation) and the extensive deployment of distributed energy assets (heterogeneity). An additional trend, coordination, is emerging as a crucial element in South Australia, facilitating the control of distributed energy assets through smart interfaces and digital networks. We then analyse the system-wide evolutions driven by the convergence of renewable energy transition and digitalization, detailing the evolving energy governance arrangements and underlining the significant interaction between public bodies and private initiatives in investments and regulatory reforms. Finally, we explore the broader implications of these developments, raising critical questions around decentralization, local empowerment, and participation in the energy transition. We conclude by suggesting that further research is essential to fully understand the complex, intertwined effects of the renewable and digital transitions
La gouvernance des plans-plages de Gironde et des Landes à l'aune des repositionnements locaux et nationaux
No full textThe current governance of beach management plans—the so-called “plans-plages”—on the Aquitaine coast (France) is the legacy of a centralized development approach, led by MIACA in the 1980’s, aimed at combining tourism appeal with the preservation of environments. Over time, it has significantly evolved due to decentralization, introducing new stakeholders to the decision-making process, and adapting to numerous local and territorial specificities. An analysis of the various types of the plans-plages in the French Gironde and Landes departments reveals a mosaic of specific situations, with varying power dynamics within the core governance partnership consisting of ONF (French National Forests Office) and municipalities or inter-municipal cooperation organizations. While the historical presence of ONF and its expertise in dune and forest management make it a key stakeholder in coastal management, local authorities are now expressing new expectations regarding the clarity of ONF actions, cost transparency, and the harmonization of contractual frameworks. Nowadays, the consequences of climate change, coupled with administrative and budgetary changes within ONF, present a significant adaptation challenge for this partnership, potentially jeopardizing the established trust between the parties. These uncertainties suggest potential strategic repositionings, both internally within ONF and among local authorities, with major implications not only for the plans-plages and their governance but also for the overall planning and management of the coastline. The present report firstly aims to analyse the current situation so as to highlight the network of stakeholders involved in the governance of the plans-plages, territorial dynamics, and ONF’s relationship with local authorities, oscillating between trust and mistrust. It then examines ongoing and future repositioning efforts, which may raise concerns among the various stakeholders. The consequences of these choices and developments are then explored and evaluated through three prospective scenarios. Ultimately, the report seeks to provide analyses and operational recommendations to enable ONF to clarify its action strategy for the plans-plages along the Aquitaine coast.La gouvernance actuelle des plans-plages du littoral aquitain est l’héritage d’un aménagement centralisé, piloté par la MIACA dans les années 1980, visant à combiner attractivité touristique et préservation des milieux naturels. Elle a, par la suite, largement évolué avec la décentralisation, introduisant de nouveaux acteurs autour de la table, et a dû s’adapter à de nombreuses spécificités locales et territoriales. L’analyse des différents types de plans-plages dans les départements de Gironde et des Landes met ainsi en évidence une mosaïque de situations particulières, avec des rapports de force variables au sein du binôme central de la gouvernance, constitué de l’ONF et des communes ou EPCI. Si la présence historique de l’ONF ainsi que son expertise sylvo-dunaire en font un acteur essentiel du littoral, de nouvelles attentes locales se font toutefois entendre du côté des collectivités en matière de lisibilité de l’action de l’ONF, de transparence des coûts et d’harmonisation du cadre contractuel. Aujourd’hui, les conséquences du changement climatique, couplées à des évolutions administratives et budgétaires de l’ONF, constituent un défi d’adaptation pour ce binôme, au sein duquel la relation de confiance établie risque de se dégrader. Les incertitudes actuelles laissent ainsi présager des repositionnements stratégiques, tant en interne pour l’ONF que du côté des collectivités locales, dont les répercussions sont importantes, non seulement sur les plans-plages et leur gouvernance, mais aussi sur l’aménagement et la gestion du trait de côte dans sa globalité. Le présent rapport vise tout d’abord à analyser la situation actuelle afin de mettre en lumière le réseau d’acteurs impliqués dans la gouvernance des plans-plages, les dynamiques territoriales et les enjeux du relationnel ONF–collectivités, entre confiance et méfiance. S’ensuit un regard sur les repositionnements en cours et à venir, qui peuvent être une source d’inquiétude pour les différents acteurs. Les conséquences de ces choix et évolutions sont ensuite explorées et évaluées à travers trois scénarios prospectifs. In fine, ce rapport ambitionne de proposer des analyses et recommandations opérationnelles afin de permettre à l’ONF une clarification de sa stratégie d’intervention sur les plans-plages du littoral aquitain