Projects
Ongoing
We develop and evaluate ML methods for fall risk prediction in nursing care. We investigate the potential of synthetic health data under strict privacy constraints. Funded by the Federal Ministry for Research and Education
ProKIP accompanies, advises and evaluates AI for nursing care research projects. Funded by the Federal Ministry for Research and Education
During heavy rainfall, polluted wastewater pollutes surface water bodies. We develop ML methods to better model sewer system dynamics.
Investigating risks and potential of synthetic satellite imagery for citizen-based monitoring. In collaboration with Alexander Glaser (Princeton Univ.) and Rebecca Frank (Univ. of Michigan)
We finetuned the Manifestoberta model of the WZB Berlin on our PoliTweets dataset to predict the political leaning of a text.
Completed
Consumers often want to shop sustainably, but sustainability information is not available or trustworthy. In collaboration with Prof. Tilman Santarius, TU Berlin ecosia we develop ML solutions to provide scalable and trustworthy sustainability information to consumers at the point of sales. Funded by the German Federal Ministry for the Environment
Heating accounts for a major part of our CO2 emissions and can be reduced by accounting for local weather conditions. We develop ML methods to forecast energy savings for public buildings in Berlin. Funded by the Berlin Senate in collaboration with Senercon.