Scientific Publications of the University of Toulouse II Le Mirail
Not a member yet
92205 research outputs found
Sort by
Trunk-and branch-data derived radial increments of Pinus brutia Ten., Cupressus sempervirens L., and Quercus pubescens Willd. subsp. pubescens from Crete for dendro-and anthraco-typological analysis
No full textInternational audienceThis dataset provides tree-ring width series for three key Mediterranean tree species that were commonly used in Aegean Bronze Age architecture: Pinus brutia Ten (Turkish pine), Cupressus sempervirens L. (Mediterranean cypress), and Quercus pubescens Willd. subsp. pubescens (Downy oak). The dataset, which was collected from well-characterised forest stands in eastern Crete, includes measurements from both trunk and branch slices. A subset of branch samples was experimentally carbonised to quantify radial shrinkage, a critical correction factor for analysing archaeological charcoal. Together, these data offer the first species-specific anatomical references for distinguishing trunk-and branch-derived wood based on tree-ring width. The dataset supports applications in dendrochronology, anthracology, and wood-use reconstruction, including the identification of selective logging or fuel management practices. It also includes R scripts for applying a generalised linear model to identify the anatomical origin of archaeological wood remains. Chronologies span up to 155 years for cypress, 130 years for pine, and 55 years for oak. While the pine and oak trunk chronologies have provided reliable crossdating, the cypress chronologies and branch datasets are more suitable for anthraco-typological analysis, i.e. for differentiating branch and trunk wood based on anatomical and growth-ring. Nevertheless, correction of the cypress series based on robust chronology will provide an interesting growth ring chronology for Crete
Finite Convergence of the Moment-SOS Hierarchy on the Product of Spheres
No full textInternational audienceWe study the polynomial optimization problem of minimizing a multihomogeneous polynomial over the product of spheres. This polynomial optimization problem models the tensor optimization problem of finding the best rank one approximation of an arbitrary tensor. We show that the moment-SOS hierarchy has finite convergence in this case, for a generic multihomogeneous objective function. To show finite convergence of the hierarchy, we use a result of Huang et al. [SIAM J. Optim. 34(4) (2024), pp 3399-3428], which relies on local optimality conditions. To prove that the local optimality conditions hold generically, we use techniques from differential geometry and Morse theory. This work generalizes the main result of Huang [Optim. Lett. 17(5) (2023), pp 1263-1270], which shows finite convergence for the case of a homogeneous polynomial over a single sphere
Des interventions de clowns pour favoriser l’adaptation au vieillissement
No full textInternational audienc
A Non-Interventionist Approach to Causal Reasoning Based on Lewisian Counterfactuals
No full textInternational audienceWe present a computationally grounded semantics for counterfactual conditionals in which i) the state in a model is decomposed into two elements: a propositional valuation and a causal base in propositional form that represents the causal information available at the state; and ii) the comparative similarity relation between states is computed from the states' two components. We show that, by means of our semantics, we can elegantly formalize the notion of actual cause without recurring to the primitive notion of intervention. Furthermore, we provide a succinct formulation of the model checking problem for a language of counterfactual conditionals in our semantics. We show that this problem is PSPACE-complete and provide a reduction of it into QBF that can be used for automatic verification of causal properties
Mise en récits de la collection cartographique du CROG : entre histoire et médiation
No full textInternational audienc
Epileptic seizures recorded with microelectrodes: A persistent multiscale gap between neuronal activity, micro-, and macro-LFP?
No full textInternational audienc
Continuity and Uniqueness for BV minimisers
No full textWe establish continuity and uniqueness results for the BV (Ω) minimisers of multidimensional scalar variational problems formulated on a bounded open set Ω. The integrand is assumed to be convex (but not necessarily strictly convex), with linear growth from below (but not necessarily from above), while the domain Ω and the boundary condition must satisfy suitable geometric and regularity conditions, such as convexity or Lipschitz continuity
De la théorie à l'application industrielle : sélection et évaluation fondées sur des principes des méthodes de tests de non-régression
No full textNational audienceThe optimization of regression testing (RT) has been widely studied in the literature, and numerous methods exist. This work focuses on the practical deployment of RT methods in Industry. It addresses the following challenge: among the many existing methods, how to determine which ones are suitable for a specific industrial context?The thesis investigates a taxonomy-based approach to assist practitioners in the characterization of their RT problem, the selection of feasible candidate solutions and their evaluation in the industrial context. The contributions are the following: 1) the proposal of an approach that builds on a recent taxonomy of RT concepts, allowing for an alignment between RT problems and candidate RT solutions; 2) an empirical assessment of the approach through a real-world case study, involving two software systems embedded into connected vehicles; 3) feedback on the usability of the taxonomy; 4) feedback on the real-world problem, including the constraints faced in the evaluation of RT methods, and the results of this evaluation. The approach delivered on its promise, enabling the identification of RT methods suitable for the studied problem. However, this success came at the cost of difficulties due to unclear taxonomy elements, missing ones, and errors in the taxonomy-based classification of methods. We conclude that the proposed alignment between problems and solution is relevant in principle, but that the taxonomy would have to mature. Our work identifies avenues for improvement.L'optimisation du test de non régression a été très étudié dans la littérature, donnant lieu à de nombreuses méthodes. Nos travaux s'intéressent au déploiement de ces méthodes dans l'industrie. Ils traitent le défi suivant : parmi les nombreuses méthodes de test de non régression, comment déterminer celles qui seront les plus adaptées à un contexte industriel spécifique ?La thèse explore une méthodologie pour aider les praticiens à caractériser leur problème de test, sélectionner les méthodes candidates applicables et lesévaluer dans le contexte industriel. Les contributions sont : 1) la définition d'une méthodologie basée sur une taxonomie récente du test de non-régression, permettant un alignement entre le problème de test et les solutions candidates ; 2) l'évaluation empirique de la méthodologie par une étude cas industrielle, impliquant deux systèmes embarqués dans des véhicules connectés; 3) des retours détaillés sur l'utilisation de la taxonomie ; 4) des retours sur l'exemple réel étudié, incluant l'analyse des contraintes rencontrées pour évaluer les méthodes candidates, ainsi que l'analyse des résultats de cette évaluation. La méthodologie a tenu ses promesses, permettant de trouver des méthodes de test de non régression adaptées au problème étudié. Cependant, ce succès a été obtenu au prix de difficultés dues à des concepts peu clairs ou manquants dans la taxonomie, et à des erreurs dans la classification des méthodes selon la taxonomie. Nous concluons que l'alignement proposé entre problèmes et solutions est pertinent dans son principe, mais que la taxonomie aurait besoin mûrir. Nos travaux identifient des pistes d'amélioration
Asymptotic stability and ergodic properties of quantum trajectories under imperfect measurement
No full textWe investigate the asymptotic stability and ergodic properties of quantum trajectories under imperfect measurement, extending previous results established for the ideal case of perfect measurement. We establish a necessary and sufficient condition ensuring the convergence of the estimated trajectory, initialized from an estimated state, to the true trajectory. This result is obtained assuming that the associated quantum channel is irreducible. Building on this, we prove the uniqueness of the invariant measure and demonstrate convergence toward this measure
What do derivational paradigms tell us about back-formation and what does back-formation tell us about derivational paradigms?
No full textInternational audienceBack-formation is the process that leads to the coinage of a lexeme by deleting a sequence present in the base, such as the verb babysit from the noun babysitter. Back-formed words exhibit a distortion between a meaning that is more complex than that of their base and a form that is simpler. The very nature of this process remains to be established. Our study suggests that back-formation (i) encompasses multiple realities, (ii) performs a formal regularization within paradigms, and (iii) most of the time involves meaning-form discrepancies. Based on five case studies of French verbs that are either back-formed from nouns or adjectives, or have the structure of neoclassical compounds, we propose that these verbs fall into three main groups, depending on the structures they involve. Our analysis highlights the importance of a paradigmatic description where formal and semantic dimensions are distinguished. We also show how a paradigmatic framework such as ParaDis sheds new light on the nature of back-formation. Conversely, back-formation helps us better characterize the structure of paradigmatic families and the relations between newly-created families and existing ones