Our mission as research center: Discover personalized medical solutions for environmentally triggered diseases to promote a healthier society in a rapidly changing world.

The Institute of Machine Learning in Biomedical Imaging (IML) focuses on research to leverage machine learning for the grand challenges in biomedical imaging in areas of unmet clinical need. Its goal is to fundamentally transform the use of imaging for diagnostics and prognostics.
The Team Reliable AI develops next-generation trustworthy artificial intelligence algorithms for medical applications. We employ advanced deep learning techniques and work on the intersection between trustworthy and probabilistic machine learning.

We are looking for a highly motivated

Postdoc in Trustworthy Artificial Intelligence (f/m/x) 101047

Image Full time
Image Neuherberg near Munich
Image Postdocs

Your responsibilities

The candidate will work at the cutting edge of deep learning research with a focus on one or more of the following topics:

  • Machine learning foundations: You will study memorisation, fairness, privacy, interpretability and model uncertainty utilising techniques from information theory and quantitative information flow. Understanding these attributes is critical for the responsible application of such models to medical settings.
  • Optimal model design for differentially private machine learning: Differential privacy (DP) is the gold-standard for privacy protection, but deep learning models trained with DP suffer from privacy-utility trade-offs. You will develop novel model architectures which leverage our increasing understanding of the behaviour of neural networks trained with DP to ameliorate these trade-offs in biomedical applications.
  • Private machine learning on correlated/graph-structured data and individual DP: The interpretation of DP guarantees in settings of correlated or graph-structured data is challenging. Moreover, it is increasingly desirable to perform individual differential privacy accounting instead of offering blanket guarantees. You will study techniques for private graph-powered and individually private deep learning for large-scale patient databases.

Your qualifications

  • Holding an excellent MSc and PhD in the field of pure or applied mathematics, (theoretical) computer science, machine learning foundations, electrical engineering, information theory, cryptography or a related field
  • A publication track record in relevant conferences (NEURIPS, ICML, ICLR, STOC) or journals including publications or works in the field of differential privacy and/or differentially private machine learning.
  • Profound knowledge of the theory of DP and machine learning foundations including advanced probability theory, information theory, statistics, real/functional analysis, etc.
  • Excellent programming and software engineering skills at least in Python and in relevant machine learning libraries (at least one of: Pytorch, TensorFlow, JAX) demonstrable through a relevant portfolio

What we offer you

work-life balance
flexible working hours & working-time models
continuous education and training
30 days annual leave
on-site health management service
Image home office options
Image on-site nursery & holiday care
Image elder care
Image company pension scheme
Image discounted public transport ticket

Munich, with its numerous lakes and its vicinity to the Alps, is considered to be one of the cities with the best quality of life worldwide. With its first-class universities and world-leading research institutions it offers an intellectually stimulating environment.

Provided that the prerequisites are fulfilled, a salary level up to E 13 is possible. Social benefits are based on the collective agreement for the federal public service (TVöD). The position is (initially) limited to two years, under certain circumstances an extension can be arranged.

To promote diversity, we welcome applications from talented people regardless of gender, cultural background, nationality, ethnicity, sexual identity, physical abilities, religion and age. Qualified applicants with physical disabilities will be given preference.
If you have obtained a university degree abroad, we require further documents from you regarding the recognition of the degree. Please request the recognition as early as possible.

If you have further questions, simply contact Sandra Mayer, 089 3187-49207, who will be happy to be of assistance.
Helmholtz Zentrum München
Deutsches Forschungszentrum für Gesundheit und Umwelt (GmbH)
Institute of Machine Learning in Biomed Imaging
Ingolstädter Landstraße 1

85764 Oberschleißheim

Back to overview Online application