Working student (f/m/x) - Machine Learning

As part of this work, you will deal with modern methods of machine and deep learning in the field of space applications. The focus is on analysing complex technical time series data and on developing, evaluating and integrating suitable methods for practical problems.
You will work on the implementation of applicable machine learning pipelines for existing simulation environments or hardware-related systems as well as on the scientific investigation of sophisticated ML methods and architectures. A particular focus is on unsupervised and semi-supervised learning methods, especially for the analysis, modelling and evaluation of time series data.

Your tasks

• Evaluation and validation of selected machine and deep learning methods using available data sets and benchmarks
• Design, implementation and further development of applicable ML pipelines for time series data
• Integration and deployment of developed solutions in existing simulation environments and/or hardware-related systems
• Scientific analysis and evaluation of advanced ML concepts, methods and architectures with regard to their applicability in the space domain
• Independent familiarisation with new scientific issues and development of relevant literature and methods
• Preparation, documentation and presentation of results
• Collaboration on scientific publications

What you bring with you

• Enrolment in a scientific Master's degree programme, preferably in computer science, mathematics, statistics, data science, aerospace or a comparable scientific and technical degree programme
• Good knowledge of at least one programming language, preferably Python
• Experience in dealing with version control and modern development processes, ideally with Git and CI/CD
• Good written and spoken English skills
• Solid knowledge of statistics as well as machine learning and deep learning
• Ideally initial experience with advanced topics such as federated learning, explainable AI, anomaly detection in time series or self-supervised learning
• Ideally knowledge or practical experience in analysing time series data
• Ideally initial experience in scientific work, for example through seminar papers, project work or theses as well as in writing scientific texts
• Ideally submit code examples, a Git repository or other evidence of practical programming experience with your application

Remuneration will be paid up to pay group 3/5 TVÖD, depending on qualifications and tasks assigned.