Computer Scientist - Interpretable Machine Learning Models (m/f/d)

Beschreibung

A job opening is available in the data science platform of the Research Group "Host Pathogen Interactions" at the next possible starting date

Artificial Intelligence / Machine Learning Expert (m/f/d)



About the Paul-Ehrlich-Institut

We are a senior federal authority responsible for vaccines and biomedicines that performs scientific research in the field of life sciences. Our institute networks with national and international players, which makes it a competent partner for the general public, science, medicine, politics and economics.

About the group

The group focuses on the understanding of virus-host interactions on a systems-level and on developing novel methodologies for analysis of large datasets (doi: 10.1038/nature08699; doi: 10.1038/s41467-018-04671-1; doi: 10.1016/j.chom.2015.11.002; doi: 10.1007/s00109-021-02131-w). The group incorporates a subgroup that addresses application of machine learning and text mining approaches to support regulators and assessors at PEI. The opening is part of a third-party funded project.Your Job Duties and Responsibilities

• Use of interpretable and eXplainable AI (XAI) methods to validate machine learning models
• Development of alternate interpretable machine learning models
• Advise third parties on the use of AI in the development and manufacture of medicinal products
• Develop guidelines and standards for the validation of artificial intelligence use in the development and manufacture of medicinal products
• Interaction and collaboration with regulators and researchers

Your Education and Experience

• A university degree (master) or Ph.D. in Computer Science or related field
• Excellent machine learning experience with a focus on XAI techniques
• Excellent knowledge of R and/or Python, or show demonstrable skill in some other programming language
• Practical experience with machine learning toolkits such as sklearn and xgboost
• Familiarity with XAI concepts such as local interpretability and Shapley values
• Familiarity with interpretation of neural networks using methods such as generative adversarial networks (GANs)
• A good knowledge of Linux and high-performance computing (HPC)
• A good knowledge of statistical methods
• Clear and effective verbal and written communication skills (English / German)
• The ability to work independently as well as tenacity and curiosity in approaching challenges
• Good communication and writing skills
• Demonstrated ability to work on multi-disciplinary teams with diverse backgrounds
• Curiosity in approaching challenges and a solution-oriented way of working

Our Working ConditionsA multifaceted and challenging position centrally located in the Rhine-Main area (Frankfurt) with good public transport connections. Being an employer in the German public service, we offer the usual social benefits
The successful candidate (m/f/d) will be employed on a fixed-term contract until December 31, 2024. Part-time working is possible in principle. The payband is E 14 TVöD Bund (Collective Bargaining Agreement for the Public Service). The assignment to a payband is made based on the consideration of your individual professional experience. Disabled candidates (m/f/d) will be given preference if their skills and qualifications are equivalent to those of non-disabled candidates. The Paul-Ehrlich-Institute encourages gender equality and life-work balance, especially in respect to the family. For this purpose, we have various employment models to offer including flexible working hours (part-time work, flexible working from home, teleworking, no core time), child care facilities and various opportunities for further professional training.