PhD position (m/f/d) in the field of Applied Machine Learning

The research aims to develop accurate, robust, interpretable, and operationally useful machine-learning methods for renewable energy and power grid forecasting, supporting the reliable operation of energy infrastructure throughout the green energy transition. The central objective is to move beyond purely accuracy-driven forecasting toward models that remain reliable under changing weather conditions, evolving energy systems, and limited or heterogeneous data. Particular focus is on: • Robust forecasting under distribution shifts • Multimodal and spatio-temporal forecasting using weather, energy and event information • Uncertainty-aware forecasting The overall goal is to develop a data-efficient and uncertainty-aware ML forecasting framework that is robust to distribution shifts, supports reliable day-ahead forecasting for renewable energy sources and power losses in electrical grids, and leads to high-impact publications supported by regular scientific supervision and mentorship.