The Data Science and its Applications (DSA) Research Group is a constellation of researchers who—while geographically and institutionally spread—are united under the supervision of Professor Sebastian Vollmer, with a shared goal of advancing data science methods and tools, and using them across industrial and socially-important applications.
Our main homes are the German Research Center for Artificial Intelligence (DFKI), where we form the Data Science and its Applications research department, and the University of Kaiserslautern, where we form the Applied Machine Learning group.
The DSA Research Group currently focuses on several major streams:
- AI for healthcare and public policy
- The research group has applied machine learning methods to develop tools to generate actionable insights to inform the British government’s response to Covid-19, perform treatment selection for diabetes and predict emergency admissions in Scotland.
- Bayesian methodology and Monte Carlo methods
- The development of novel methodologies and the extension of established theoretical works pertaining to machine learning.
- Event analysis
- The interconnected problems of measuring and predicting individual and societal health and well-being, through the use of longitudinal surveys, time series and survival data to model both directly observable and latent temporal states, such as adverse health events.
- Machine Learning in Julia
- Development of the software package Machine Learning in Julia (MLJ), a modelling toolbox providing a common interface and meta-algorithms for selecting, tuning, evaluating and building composite models.
- Responsible AI
- Modern technologies introduce novel challenges in algorithmic fairness, data privacy, and scientific reproduciblity. We are developing methods to detect and mitigate ethical issues in AI-assisted processes.
- Data Science for Social Good
- We use our knowledge in these areas to champion new ways of delivering impact, through short term collaborations with industrial, nonprofit and academic partners to address real-world challenges in programs that combine training and delivery.
Latest News
On March 18, 2022, German Chancellor Olaf Scholz visited the DFKI site in Kaiserslautern together with the Minister President of Rhineland-Palatinate, Malu Dreyer.
On Wednesday the 19th of January, 2022, Prof Sebastian Vollmer was acting as a witness in an evidence session of the UK House of Commons Science and Technology Committee on reproducibility and research integrity
I have received many questions about survival neural networks (‘survival networks’) in R, ranging from “is this even possible?” to “how do I install Python in R?” and “how can I tune these models?”. If you are an R user with an interest in survival networks then this is the article for you!
Short account of Harry’s internship
MLJ (Machine Learning in Julia) is a toolbox written in Julia providing a common interface and meta-algorithms for selecting, tuning, evaluating, composing and comparing machine learning models written in Julia and other languages.
