The research at MCML in Medicine and Healthcare focus on objectives that are necessary to overcome the hurdles for the deployment of ML approaches in clinical environments. In particular, advances are required in interpretable and explainable deep learning, robust and data efficient learning, privacy preserving learning, as well as in trust and safety of autonomous AI and ML systems.
MCML focuses on crucial issues in Biology and Biomedicine, addressing AI challenges such as liability, black-box behavior, and privacy. The goals include advancing personalized healthcare and fostering collaboration between algorithms and human experts. Additionally, MCML aims to be a key training hub for the next generation of AI-empowered professionals in medical and biological fields.
Geoinformation from Earth Observation satellite data is vital for addressing societal challenges. The research focus at MCML in this area is on tailoring data science and ML for geo-relevant applications. This includes physics-aware ML, uncertainty quantification, explainable geoinformation retrieval, Quantum ML for a digital twin of the Earth, and ethical considerations in ML for Earth Observation.
The landscape for empirical social scientists has transformed with the rise of computational social science. Our researchers focus on aligning research goals with available digital trace data, evaluating data quality in relation to research objectives, and ensuring reproducibility through thorough documentation. They emphasize the critical need to assess and evaluate data feeding into AI systems to prevent biases, unfair operations, and the exacerbation of social inequalities.
MCML emphasizes ML and data science research for human benefit, improving actions, automating tasks, and revealing insights. Basic ML research, though generic, offers wide applicability. In human-centered ML, we prioritize efficient human-algorithm-data interaction, expanding beyond traditional human-computer interaction to include intelligent systems and data, all within a framework of ethical considerations in AI development and deployment.