Inicio arrow Becas y Ofertas de Empleo arrow Three PhD Positions on the Intersection of Formal Methods, Artificial Intelligence, and Machine Lear
Three PhD Positions on the Intersection of Formal Methods, Artificial Intelligence, and Machine Lear
martes, 02 de junio de 2020

The overall challenge in the NWA-funded PrimaVera project is to fulfil the grand promises in predictive maintenance. Through an effective combination of sensor techniques, big data analysis, and maintenance engineering, we want to significantly improve failure predictions to render maintenance more effective. You will work on the intersection of formal methods (such as fault trees, stochastic model checking) and data analytics (decision trees, Bayesian Networks, neural networks, reinforcement learning, POMDPs). PrimaVera is a joint project with Eindhoven University of Technology, the University of Twente, Saxion, Hague University of Applied Sciences and the Dutch Aerospace Laboratory, as well as several industrial partners. You will be expected to participate in a fruitful collaboration with the industrial partners, for example, by carrying out an industrial case study. The project will be supervised by Dr. Nils Jansen and Prof. Marielle Stoelinga.

* Provably Correct Policies for Uncertain Partially Observable Markov Decision Processes
This NWO-funded project is dedicated to safety-critical artificial intelligence (AI) scenarios that may suffer from data uncertainty. In particular, we will bridge the gap between AI and formal methods through models that incorporate incomplete information and uncertain outcomes of decisions, namely partially observable Markov decision processes (POMDPs). The two central research goals are: (1) to develop scalable methods that guarantee decision making with hard guarantees on safety constraints under uncertainty, and (2) to render (deep) reinforcement learning to adhere to specifications during the exploration of uncertain and partially unknown POMDPs. The project is purely academic, but involves potential collaboration and research visits with well-known researchers from all over the world. The project will be supervised by Dr. Nils Jansen.

* SAM-FMS - Scheduling Adaptive Modular Flexible Manufacturing Systems
SAM-FMS is an NWO-funded multi-partner project together with Eindhoven University of Technology, Delft University of Technology, ESI (TNO), and Canon Production Printing. The project will deliver new scheduling and co-design methods by pursuing model-driven synthesis approaches based on flow-shop models of manufacturing systems. We will employ and develop techniques from machine learning whose dependability is increased through formal methods and formal verification. You will work closely with the internal and external partners to help achieve the SAM-FMS project objectives. Specifically, close collaboration is needed with the industrial partners, the Electronic Systems group at Eindhoven University of Technology, and the Algorithmics group at Delft University of Technology. The project will be supervised by Dr. Nils Jansen and Prof. Frits Vaandrager.

The positions are available within the Institute for Computing and Information Sciences (iCIS) at the Faculty of Science of Radboud University Nijmegen. Research at iCIS focuses on software science, digital security and data science. During recent evaluations, iCIS has been consistently ranked as the No. 1 Computing Science department in the Netherlands. The Software Science group is well known for its contributions to the mathematical foundations of software, formal methods, AI, machine learning, and functional programming, with strong ties to international top researchers.

If you are interested, please *exclusively* apply via the following webpage where you can also find more information:

For further questions, please contact Nils Jansen< Esta dirección de correo electrónico está protegida contra los robots de spam, necesita tener Javascript activado para poder verla >

Further information:

Dr. Nils Jansen
Assistant Professor
Department of Software Science
Radboud University Nijmegen