Closing date 10 August 2018
The Informatics Institute (IvI), one of the large research institutes of the Faculty of Science, with a focus on complex information systems divided in four broad themes: Artificial Intelligence, Computational Science, Data Science en Systems and Network Engineering. They have a prominent international standing and are active in a dynamic scientific area, with a strong innovative character and an extensive portfolio of externally funded projects.
A PhD position opening is now available on surrogate modelling for multiscale simulations.
Project description
In Multiscale Modelling a macroscale model is typically coupled to a few microscale models, where the microscale models are usually the ones that require very large computing resources. Examples could be coupling continuous models of materials or fluids to molecular of mesoscopic models that would capture the micro dynamics of e.g. a complex material or fluid. In this project classes of multiscale models are considered where the amount of executions of the microscale models becomes so exceedingly high that, even on the fastest supercomputers, simulation times would become prohibitive. The goal of this project is to develop advanced meta modelling techniques that would render such applications tractable. The main focus will be on two scenarios. The first is on Uncertainty Quantification of multiscale models, the second on Heterogeneous Multiscale Models (HMM).
The research will be driven forward by applications within the Computational Science Lab, mainly from the biomedical domain, and in relation to the development of the Virtual Artery. The research will be embedded within a large European collaboration called VECMA, where nine laboratories throughout Europe collaborate to enable a diverse set of multiscale applications (from fusion and advanced materials through climate and drug discovery to clinical decision making in personalized medicine) to run on current multi-petascale supercomputers such that the simulations are Validated, Verified, and equipped with Uncertainty Quantification (UQ), and which will deliver a VVUQ toolkit for multiscale computing on High Performance Computers.
VECMA is funded by the European Commission, where 9 laboratories from all over Europe collaborate. Within the Netherlands the University of Amsterdam and the Center for Mathematics and Informatics (CWI) participate in VECMA. As PhD student, you will be in the unique position to collaborate intensively with the other labs that participate within VECMA. You will have regular contacts with the Dutch partner and will be encouraged to spent part of your time in other labs within the VECMA project, e.g. in London or Munich.
The candidate will also collaborate with the eMUSC project, a collaboration between the University of Amsterdam and the Dutch eScience center, where algorithms and libraries for UQ for multiscale applications are developed.
If you're interested to apply further information can be found here
