Ufuk is a DPhil student and part of the Intelligent Earth CDT in AI for the Environment. His research focuses on decision making under uncertainty for wildfire management, investigating how reinforcement learning can support human decision makers in responding to wildfires while remaining ethical and safe.
Prior to his DPhil, he received a Master of Science in Physics at Heidelberg University, where he worked on machine learning applications in computational astrophysics. The 2023 earthquake in Turkey and Syria motivated his shift toward understanding what "better" decisions really mean under high stakes and deep uncertainty in hazard scenarios.
Most Recent Publications
JaxWildfire: A GPU-Accelerated Wildfire Simulator for Reinforcement Learning
JaxWildfire: A GPU-Accelerated Wildfire Simulator for Reinforcement Learning
Neural Value Iteration
Neural Value Iteration
RUBIX: Differentiable forward modelling of galaxy spectral data cubes for gradient-based parameter estimation
RUBIX: Differentiable forward modelling of galaxy spectral data cubes for gradient-based parameter estimation
Fast GPU-Powered and Auto-Differentiable Forward Modeling of IFU Data Cubes
Fast GPU-Powered and Auto-Differentiable Forward Modeling of IFU Data Cubes
MEGS: Morphological Evaluation of Galactic Structure
MEGS: Morphological Evaluation of Galactic Structure