Scientific Publications of the University of Toulouse II Le Mirail
Not a member yet
92205 research outputs found
Sort by
Rate of Convergence in the Functional Central Limit Theorem for Stable Processes
No full textInternational audienceIn this article, we quantify the functional convergence of the rescaled random walk with heavy tails to a stable process.This generalizes the Generalized Central Limit Theorem for stable random variables infinite dimension. We show that provided we have a control between the randomwalk or the limiting stable process and their respective affine interpolation, we canlift the rate of convergence obtained for multivariate distributions to a rateof convergence in some functional spaces
Enregistrement numérique multipistes et immersion : l’équipe son du plateau à laconquête de l’espace sonore ?
No full textInternational audienc
Low dimensional representation of multi-patient flow cytometry datasets using optimal transport for minimal residual disease detection in leukemia
No full textInternational audienceRepresenting and quantifying Minimal Residual Disease (MRD) in Acute Myeloid Leukemia (AML), a type of cancer that affects the blood and bone marrow, is essential in the prognosis and follow-up of AML patients. As traditional cytological analysis cannot detect leukemia cells below 5\%, the analysis of flow cytometry dataset is expected to provide more reliable results. In this paper, we explore statistical learning methods based on optimal transport (OT) to achieve a relevant low-dimensional representation of multi-patient flow cytometry measurements (FCM) datasets considered as high-dimensional probability distributions. Using the framework of OT, we justify the use of the K-means algorithm for dimensionality reduction of multiple large-scale point clouds through mean measure quantization by merging all the data into a single point cloud. After this quantization step, the visualization of the intra and inter-patients FCM variability is carried out by embedding low-dimensional quantized probability measures into a linear space using either Wasserstein Principal Component Analysis (PCA) through linearized OT or log-ratio PCA of compositional data. Using a publicly available FCM dataset and a FCM dataset from Bordeaux University Hospital, we demonstrate the benefits of our approach over the popular kernel mean embedding technique for statistical learning from multiple high-dimensional probability distributions. We also highlight the usefulness of our methodology for low-dimensional projection and clustering patient measurements according to their level of MRD in AML from FCM. In particular, our OT-based approach allows a relevant and informative two-dimensional representation of the results of the FlowSom algorithm, a state-of-the-art method for the detection of MRD in AML using multi-patient FCM
The role of plant substances in the ancient diet of Corsica during Protohistory
No full textInternational audienc
Le paysage ligneux des populations littorales de Corse au Néolithique ancien : données récentes de l’anthracologie des sites de Basí, A Petra et Araguina-Senola
No full textInternational audienc
Discovering search behaviour in black garden ant trajectories
No full textExploration of space plays an important role in many animals, in particular for social insects who have to feed and protect a whole colony. In a laboratory study, Khuong et al (2013) studied how the workers of the black garden ant Lasius niger move around in an unknown environment. They assumed that, in a homogeneous arena with no visual cues, ants had no information about their position in space. Based on this hypothesis, they modelled their ants in a Boltzmann Walker framework which describes an ant's random walk as a series of straight segments separated by reorientation events. They assumed that on plain horizontal surfaces the ant's direction of movement would not influence their average speed, segment lengths and reorientation decisions, thus leading to diffusive trajectories. However, published experiments indicate that L. niger ants are not completely devoid of directional information even in standard laboratory setups with no obvious landmarks. Moreover, many ant species are known to develop specific search strategies when they want to find a particular place in space, a situation that may apply to the analysed data. We re-analyze Khuong et al ’s data on non-inclined surfaces, this timetaking into account the ant’s orientation in relation to its starting point in the arena. We discovered that, with this information taken into account, the ant’s trajectory is biased towards its starting point (biased random walk), revealing an advecto-diffusive process. In fact, the distributions of segment lengths and reorientation angles turned out to be modulated by the ant’s orientation in relation to its starting point. By simulating these biased trajectories, we show that this modulation halves the time it takes for an ant to come back towards its starting point. We conclude that not taking into account the animal’s cognitive abilities in data analysis may lead to incomplete or biased conclusions. The discovered search behaviour in L. niger can play a significant role in the colony’s exploration and foraging ecology
Plant-Pollinator Interactions in the Anthropocene: Why We Need a Systems Approach
No full textInternational audienceSynopsis Animal-mediated pollination is one of the most ecologically and economically important mutualisms and serves as a remarkable example of cross-kingdom communication and coevolution. Unfortunately, pollinators, plants, and the interactions between them are threatened in the Anthropocene. While pollination emerges from interactions across biological scales, existing research and expertise have developed in distinct silos reflecting traditional fields of study such as ecology, plant physiology, neuroethology, etc. This forward-looking review and perspective is a culmination of the “Plant-pollinator interactions in the Anthropocene” symposium at the 2025 Society for Integrative and Comparative Biology meeting, which collected expertise across these disciplinary silos to identify pressing questions our community needs to tackle in the next decade. In this perspective piece, we argue that an integrative, organismally informed systems approach is critical to unraveling the complexity of how plant-pollinator relationships are impacted by dynamic anthropogenic stressors. Specifically, this calls for an intentional and iterative integration of holistic modeling studies with empirical studies. Modeling the emergent properties driven by organismal interactions in pollination systems can identify impactful variables; this in turn should drive design of empirical studies that elucidate how organisms respond to changing environments in the context of those impactful variables, feeding back into improved models. Repetition of this process will allow better predictive power over pollination stability in changing landscapes. Finally, we consider both existing barriers to this integration, as well as emerging opportunities (such as new technologies) that can help bridge across traditional fields
Social attraction mediates collective foraging decisions in invasive hornets
No full textGroup-living animals commonly use social information to better locate and exploit resources. In many insects, birds, fish and mammals, this can lead to collective foraging decisions by which animals share a single food source among alternatives of equal qualities. Here, we report collective foraging decisions in a social wasp, the yellow-legged hornet Vespa velutina nigrithorax, a major predator of bees and other terrestrial invertebrates invasive across Asia, Europe and North America. When given a choice between two identical liquid food sources (feeders or traps containing sugar solutions), wild hornets distributed asymmetrically on the two options, and this phenomenon was more frequent as group size increased. Priming one of the food sources with dead hornets predictably biased the collective choices towards this particular option, irrespective of whether the dead insects were conspecifics or hornets from a closely related species. Inter-attraction in yellow-legged hornets is thus a passive and non-specific mechanism, possibly mediated by visual or chemical cues displayed by dead hornets. This collective behaviour may provide important foraging advantages to hornets invading new territories and bring new perspectives for population control
e-FRAN-Transfert: Des orientations scientifiquement consolidées
No full textProgress report on the 9 projects of the e-FRAN-Transfer action included in France 2030, led by Mission Monteil with ANR as operator.Rapport d'étape des 9 projets de l'action e-FRAN-Transfert inscrite dans France 2030 pilotés par la Mission Monteil avec l'ANR comme opérateur
Regional stability conditions for recurrent neural network-based control systems
No full textInternational audienceIn this paper we propose novel global and regional stability analysis conditions based on linear matrix inequalities for a general class of recurrent neural networks. These conditions can be also used for state-feedback control design and a suitable optimization problem enforcing norm minimization properties is defined. The theoretical results are corroborated by numerical simulations, showing the advantages and limitations of the methods presented herein