-
barriers: a large input modality gap, as network data consists of diverse, non-textual formats like time-series metrics, graphs, and scalar values; the inefficiency and unreliability of answer generation
-
such as Alzheimer's - but predicting how proteins fold in biological environments remains a key unmet challenge. This project brings together insights from efficient graph-driven folding simulations with
-
such as graph-based approaches and network analytics to predict how blue network dynamics, fragmentation and surrounding land use interact to shape ecosystem functioning and human wellbeing outcomes
-
the connections between clouds and climate. Ultimately, we want to create to causal graphs for large-scale cloudiness, its dependence, and its effect on the related environmental factors. Additional or alternative
-
graph, and discrete random processes. The aim of this project is for the student to develop an understanding of these tools and to apply these techniques to open research problems in the field. Entry
-
including: * Algorithmic game theory * Approximation algorithms * Automata and formal languages * Combinatorics and graph algorithms * Computational complexity * Logic and games * Online and dynamic
-
learnable physical kernels, geometry encodings, and boundary-aware layers; compare to PINNs, U-Nets, graph operators, and transformer baselines. Learning strategy: physics-informed and multi-task losses
-
be developing advanced spatial models such as graph-based approaches and network analytics to predict how blue network dynamics, fragmentation and surrounding land use interact to shape ecosystem
-
areas, including generative modelling (e.g. diffusion models, flow matching, self-supervised and autoregressive approaches), causal machine learning, graph neural networks, dynamical systems modelling
-
, storm activity, and other hazards using graph-based clustering, fuzzy machine learning, and reduced-order models – delivering scientific insight into where and when rerouting is needed. Real-Time Decision