- 
                
                
                processes that produce energy and raw materials. The Department of Thermodynamics of Actinides is looking for a PhD Student (f/m/d) - Machine Learning for Modelling Complex Geochemical Systems. The job 
- 
                
                
                of neural hydrology, where hydrological models are directly learned from data via machine learning (e.g., LSTM neural networks, [1]). Initially, these models ignored all physical background knowledge and did 
- 
                
                
                , the details of the process are not yet fully understood. Mechanistic learning, the combination of mathematical mechanistic modelling and machine learning, enables a data-driven investigation of the processes 
- 
                
                
                modeling and computational workflows Knowledge about machine learning: statistics and deep learning Experience in data analysis, visualization and presentation Good programming skills in languages such as 
- 
                
                
                -scale modelling, machine learning) High resolution analysis, monitoring of chemistry, structure and transformations at the atomic scale of buried interfaces and defects by correlated experimental 
- 
                
                
                machine learning We offer: Academic freedom to pursue your scientific interests related to infection biology, inflammation, gene expression, and intracellular organization Competitive salary including 
- 
                
                
                Mathematics/ Approximation Theory to be filled by the earliest possible starting date. The Chair of Applied Mathematics, headed by Prof. Marcel Oliver, is part of the Mathematical Institute for Machine Learning 
- 
                
                
                machine learning approaches. These are similar to earlier work on charge and excitation energy transfer (see https://constructor.university/comp_phys). The project for the PhD fellowship is slightly more 
- 
                
                
                and data analytics (including machine learning and deep learning); from high-performance computing to high-performance analytics; from data integration to data-related topics such as uncertainty 
- 
                
                
                direction (possibly forming spontaneous “lanes”), crossing, and opposite flows. For single-lane vehicular traffic, the model should revert to a car-following model. In cooperation with the supervisor Dr