The Machine Learning for Simulation Science group is working on core machine learning research with a focus on graph neural networks and geometric deep learning, the synthesis of discrete (algorithmic) components and continuous learning systems, and the intersection of machine learning and the simulation sciences.
We are looking for a HiWi to help with our research on neural surrogate models for solving partial differential equations (PDEs). PDEs play an essential role in many areas of science and are traditionally solved using numerical solvers, which can be slow and require a lot of computational resources. Thus, the goal is to replace such expensive numerical solvers with fast neural surrogate models. In our current project, we are focusing on improving the data generation phase for the neural PDE solver by using active learning techniques, i.e., by selecting the most informative data points iteratively to increase the data efficiency.
Tasks
Extend our existing research code base by incorporating new modules and algorithms.
Conduct experiments on our GPU cluster.
Proficiency in Python programming
Basic understanding of Machine Learning
Basic knowledge of PyTorch (or another ML framework).
Preferably experience with executing experiments on a GPU cluster
Competency in English or German
Enrolled as a student at the University of Stuttgart
Since we are looking to hire someone long-term, you should have some time in your studies left (i.e., be in the middle stage of a bachelor's program up to the beginning of a master's program)
We offer :
Hands-on experience in machine learning research
Familiarize yourself with the current research trends in ML for Science
Flexible work hours
Possibility to follow up with a thesis
Up to 40 hours per month contract (initially limited to 6 months, can be extended afterwards)
We are looking forward to your application! We will review them on a rolling basis until the 30.06.24. Please send a short cover letter, transcript, and CV to the contact email informed below.