Monitoring air pollution plumes in emergency situations (industrial accidents, natural disasters, deliberate terrorist releases, etc.) becomes an issue of utmost importance in cities worldwide, given the dramatic effects that the released pollutants can cause. In these situations, the pollution plume is strongly dynamic leading to a fast dispersion...
In recent years, the vehicular field has undergone significant advancements with the development of autonomous vehicles and smart cities. These advancements have brought about a modernization of human life, where everything is interconnected - from individuals through smartphones to infrastructure, cars, and motorcycles. In such a system, informati...
International audience
The intelligence of a smart home is realized by creating various services. Eachservice tries to adjust one monitored state by controlling related actuators after consideringenvironment states detected by sensors. However, the design of the logic of the services deployedin a smart home faces limitations of either dynamic adaptability (predefined rul...
Cognitive Control is the general capacity of an organism to use top-down control signals to inhibit the dominant behavior in favor of a contextually relevant response, in accordance with internally described goals (which could result from environmental or motivational factors). Various experimental studies and computational models have tried to und...
L'apprentissage profond par renforcement utilise des simulateurs comme oracles abstraits pour interagir avec l'environnement. Dans les domaines continus des systèmes robotiques multi-corps, des simulateurs différentiables ont récemment été proposés mais sont encore sous-utilisés, même si nous avons les connaissances nécessaires pour leur faire prod...
Deep reinforcement learning uses simulators as abstract oracles to interact with the environment. In continuous domains of multi-body robotic systems, differentiable simulators have recently been proposed, still, they are yet underutilized, even though we have the knowledge to make them produce richer information. This problem when juxtaposed with ...
Cette thèse porte sur l'amélioration des campagnes enchères pour l'affichage publicitaire en ligne. Nous considérons le problème à travers deux grandes questions : la prédiction de la probabilité de clic permettant d'obtenir une estimation de la valeur d'un affichage, et l'optimisation des enchères qui doit, en se basant sur cette estimation, gérer...
The experience gained by interacting with the surrounding world is the main mean by which animals and humans learn. The mammalian brain can re-elaborate past experiences and contextually organize them, through neural circuitries which involve the hippocampus. Hippocampal reactivations of place cells seem to exploit experience to infer the outcome o...
Markovian bandits are a subclass of multi-armed bandit problems where one has to activate a set of arms at each decision time. The activated arms evolve in an active Markovian manner. Those that are not activated (i.e., are passive) either remain frozen – then one falls into the category of rested Markovian bandits – or evolve in a passive Markovia...
