Neural networks are extremely versatile tools, as evidenced by their widespread adoption into many fields in the sciences and beyond, including the geosciences. In seismology neural networks have been primarily used to automatically detect and discriminate seismic signals within time-series data, as well as provide location estimates for their sources. However, as neural network research has significantly progressed over the past three decades, so too have its applications in seismology. Such applications now include earthquake early warning systems based on smartphone data collected from large numbers of users, the prediction of peak ground acceleration from earthquake source parameters, the efficient computation of synthetic seismograms, providing probabilistic estimates of solutions to geophysical inverse problems, and many others. This thesis contains three components, each of which explore novel uses of neural networks in seismology. In the first component, a previously established earthquake detection and location method is supplemented with a neural network in order to automate the detection process. The detection procedure is then applied to a large volume of seismic data. In addition to automating the detection process, the neural network removes the need for several user-defined thresholds, subjective criteria otherwise necessary for the method. In the second component, a novel approach is developed for inverting seafloor compliance data recorded by ocean-bottom seismometers for the shallow shear-wave velocity structure of oceanic tectonic plates. The approach makes use of mixture density networks, a type of neural network designed to provide probabilistic estimates of solutions to inverse problems, something that standard neural networks are incapable of. In the final component of this thesis, the mixture density network approach to compliance inversion is applied to a group of ocean-bottom seismometers deployed along the continental shelf of the Cascadia Subduction Zone in order to investigate shelf sediment properties.

Dans la nécessité de développer des véhicules sûrs, confortables, économiques et à faible impact environnemental, les voitures sont de plus en plus équipées d'organes qui emploient des capteurs, actionneurs et systèmes de commande automatiques.Or ces systèmes, critiques pour la sécurité et le confort des passagers, peuvent mal-fonctionner en présence d'une défaillance (défaut).Dans le cadre du diagnostic à bord, plusieurs approches à base de modèle sont développées dans ce travail afin de détecter, localiser et estimer un défaut capteur ou actionneur, et pour détecter la perte de stabilité du véhicule.Ces méthodes reposent sur une synthèse robuste pour les systèmes incertains à commutation.Elles sont validées en simulation avec le logiciel CarSim, et sur les données réelles de véhicule dans le cadre du projet INOVE. / Modern vehicles are increasingly equipped with new mechanisms to improve safety, comfort and ecological impact. These active systems employ sensors, actuators and automatic control systems. However, in case of failure of one these components, the consequences for the vehicle and the passengers safety could be dramatic. In order to ensure a higher level of reliability within on board diagnosis, new methodologies for sensor or actuator fault detection, location and estimation are proposed. These model based approaches are extended for robust synthesis for switched uncertain systems. In addition, a method for detecting critical stability situation is presented. The validation of the different methods is illustrated with simulations using CarSim, and application on real vehicle data within the INOVE project.