Skip to content Skip to footer
News

PhD student Position: Physics Informed Neural Networks, Graph Neural Networks and Isogeometric Analysis Simulators

PhD student Position: Physics Informed Neural Networks, Graph Neural Networks and Isogeometric Analysis Simulators

In the framework of an exciting research grant Real-time inversion using self-explainable deep learning driven by expert knowledge (IN-DEEP) at the Faculty of Computer Science, AGH University of Kraków, Poland, led at AGH by Prof. Maciej Paszyński, we are looking for a Ph.D. student in computational science/applied mathematics to perform research at AGH University and obtain a double PhD at AGH University of Kraków, Poland, and The University of the Basque Country, Bilbao, Spain.

Job description:

The research project is related to the development of Physics Informed Neural Networks (PINN), Graph Neural Network (GNN), and isogeometric analysis (IGA) simulators. The prototype solvers will be applied to model flow in porous-fractured geological formations and to the model tumor growth simulators. Possible applications include, e.g. geothermal flow, CO2 sequestration problems, and tumor treatment.

This is an exciting research project involving the investigation of PINN and GNN, comparison with IGA simulators, and investigation of the application of PINN /GNN for the solution of parametric PDEs.

The successful candidate will join a 3-year research project focused on developing novel Neural Network based simulators for challenging applications.

We will provide an open-source IGA-ADS environment for developing IGA simulations [1].

We will also provide an open-source Python environment for running PINN simulations [2].

The prospective Ph.D. student should have:

● a Master degree from a recognized university either in computer science, mathematics, mechanical engineering, or related disciplines

● a strong background in scientific computing with C++ and Python

● experience with Neural Networks training using e.g., pytorch / tensor flow / JAX

● creativity and motivation for addressing computational science problems

● excellent English language skills

● experience with Physics Informed Neural Networks and IsoGeometric Analysis are a plus

We offer:

● a three-year contract, with possible extension

● a young and dynamic work environment at one of the top universities in Poland

● a very competitive salary (total for three years): Living allowance: 86292 euro,

 Mobility allowance: 21600 euro, Family allowance: 17820 euro.

 [the cost of a good living for a student in Krakow is around 1000 euros per month].

● being part of the Doctoral Network EU project IN-DEEP involving collaboration with universities and companies from Spain, Italy and France

Information & contact:

The project will start between March and August 2024, and applications are welcome

until the position will be filled. To submit your application, please write to

paszynsk@agh.edu.pl

 

[1] github.com/marcinlos/iga-ads

[2] github.com/pmaczuga/pinn-notebooks

Stopka