Postdoc position in Machine Learning for Advanced Physical Simulation at Aarhus University, Denmark

I have an opening in my group for a 2-year postdoc position offering applicants an exciting opportunity to join my ERC-funded project “ALPS - AI-based Learning for Physical Simulation”.

Job description

You will be contributing to developing novel algorithms at the intersection of computational physics and machine learning for the automatic identification of symbolic models of physical systems starting from experimental data.

You will be working primarily with evolutionary algorithms, neural networks, reinforcement learning, computer graphics and handling of mesh data and high-performance scientific computing libraries.

The position will focus on aspects of implementation, optimization/parallelization of the open-source library that we are developing as a part of the project for symbolic learning of physical models.

Your profile

Applicants should hold a PhD in Computer Science, Computer Engineering, Artificial Intelligence, Physics, Mathematical Engineering, Mechanical Engineering or similar.

Relevant skills

Strong Python programming skills (as a plus: C++ or Julia)

High-performance computing expertise (including use of multiprocessing libraries and GPU computing)

Deep knowledge of common neural network architectures (as a plus: architectures for NLP, such as transformer networks)

Use of Scientific Computing libraries (e.g. numpy, scipy, JAX)

Use of Machine Learning and Reinforcement Learning libraries (e.g. pytorch, Haiku/Flax, RLax, Stable Baseline)

Deadline: 15 February 2024

Apply here