1 Postdoc (m/f/d) in Deep Learning / Reinforcement Learning / CPN

Dife; Potsdam; Hybrid; EUR 60.000 - 80.000; Vollzeit

Benefits dieser Stelle

30 days vacation

Jobticket

Company pension scheme

Audit berufundfamilie

Zusammenfassung

DIfE in Potsdam-Rehbruecke is seeking a highly motivated scientist to join the Department of Computational Precision Nutrition. The role focuses on deep learning, multimodal data, and reinforcement learning to develop digital twins and personalized health recommendations.

You will collaborate across disciplines, implement models for multimodal studies, and communicate results through visuals and reports for scientific publications and participants.

Qualifikationen

• Master and doctoral degree in a relevant field.
• Publication record at top machine learning conferences.
• Proficiency in Python or R.
• Experience with advanced deep learning frameworks and open-source software development.
• Strong communication skills and ability to work in interdisciplinary teams.

Aufgaben

• Develop, implement and apply digital twins and RL agents based on multimodal health data for personalized recommendations.
• Collaborate with causal inference researchers to link multimodal data and deep learning.
• Develop methodology for individual-level inference of large epidemiological studies with omics data.
• Implement models on the StudyU platform with colleagues in multiple countries.
• Create clear visualizations, reports and summaries for publications and study participants.
• Collaborate with clinicians, epidemiologists, software developers and lab scientists on study design and analysis.

Kenntnisse

Deep Learning

Multimodal Learning

Reinforcement Learning

Python

R

Open-source software

Ausbildung

Master's degree

Doctoral degree

Tools

PyTorch

TensorFlow

Jobbeschreibung

The German Institute of Human Nutrition Potsdam-Rehbruecke (DIfE) is a member of the Leibniz Association. The institute’s mission is to conduct experimental and clinical research in the field of nutrition and health, with the aim of understanding the molecular basis of nutrition-dependent diseases, and of developing new strategies for treatment and prevention.

starting as soon as possible.

The Department of Computational Precision Nutrition develops methods and software for the analysis of both population-level and individual-level data, to enable personalized health recommendations based on dietary patterns, behaviors, and diet-associated biomarkers. We aim to contribute to the personalized prevention and treatment of chronic diseases and advance our understanding of the mechanisms underlying their development. Two methodological focus areas are digital N-of-1 trials and deep learning-based modeling of multimodal biomedical data.

We are seeking one highly motivated scientist with expertise in Deep Learning / Reinforcement Learning to join our team.

Tasks include

• Develop, implement and apply digital twins and RL agents based on multimodal health data for personalized recommendations of dietary and other health behavior to improve health
• Collaborate with causal inference researchers to jointly develop methods for analyzing multimodal digital N-of-1 trials (patient reported outcomes, wearables, images, audio, omics data) linking causal inference and deep learning
• Develop methodology for individual-level inference of large epidemiological studies (e. g. EPIC Potsdam study, German National Cohort study) including omics data
• Implement the developed models for multimodal studies run on the StudyU platform with collaborators in Germany, Australia, USA, South Korea and Ghana
• Develop clear data visualizations, reports, and written summaries to communicate results – for scientific publications but also for study participants and patients
• Collaborate with clinicians, epidemiologists, software developers and laboratory scientists in the design of new studies and analysis of existing data
  

Skills and requirements

• Excellent master and doctoral degree with demonstrated expertise in Deep Learning / Multimodal Learning / Reinforcement Learning
• Publication record at top machine learning conferences
• Expertise in programming languages such as R or Python
• Experience with advanced deep learning frameworks & open-source software development
• Strong communication skills and ability to work in interdisciplinary teams
  

We offer

• Opportunity to develop your own research profile among the exciting research topics described above
• A dynamic, international and interdisciplinary research environment as well as excellent working conditions and outstanding technical equipment
• Employment with remuneration according to TV-L, level 13, plus annual special payment and company pension scheme
• Family-friendly working conditions (certificate “audit berufundfamilie”)
• Supporting of mobility with a jobticket for using the public transport
• Location close to the vibrant city of Berlin, with easy accessibility by public transport or car
• 30 days of vacation
• Participation in the benefits program for employees („Corporate Benefits“)
  

The advertised positions are available for initially 3 years.

We promote the employment of people with severe disabilities and are committed to equal opportunities for them. Applicants with severe disabilities will be given preferential consideration if they have the same qualifications.

Contact for further information

Data privacy

Boosting Health by Nutrition Research.

DIfE is a member of the
Leibniz Association

EUROPEAN UNION
European Regional
Development Fund