Machine Discovery of Physical Models and Laws University of Bristol in United Kingdom

This PhD project at the University of Bristol's School of Engineering Mathematics and Technology explores the intersection of dynamical systems, physical modelling, and machine learning. The research aims to develop innovative methods for discovering mathematical models of complex systems that are challenging to describe from first principles. Examples of such systems include fuel sloshing inside airplane wings, jointed mechanical structures, and building vibrations during earthquakes.

Throughout your doctoral studies, you will focus on extracting minimal and interpretable models from data, enabling easier analysis and simulation of physical phenomena. The project leverages advanced mathematical and computational techniques, including invariant foliations (related to Meta's Joint Encoding Predictive Architecture, JEPA), deep neural networks, and compressed tensors or tensor networks. These approaches are at the forefront of data-driven modelling and are considered promising pathways toward human-like machine intelligence, as they facilitate conceptual understanding from unstructured data without supervision.

The research environment at the University of Bristol offers access to cutting-edge facilities and a vibrant academic community. You will be part of a multidisciplinary team, collaborating with experts in applied mathematics, computational physics, structural mechanics, and data science. The project is ideal for candidates with strong analytical skills and a passion for mathematical modelling, machine learning, and engineering applications.

Eligibility requirements include a first-class or upper second-class degree in engineering, mathematics, physics, computer science, or a closely related field. Applicants should demonstrate proficiency in analytical and mathematical methods, with experience in machine learning, computational modelling, or data analysis considered advantageous. Non-native English speakers must provide evidence of language proficiency, such as IELTS or an equivalent qualification.

The application deadline is January 31, 2026. Interested candidates should review the full project details and submit their application online via the provided link. Prepare your CV, academic transcripts, and a statement of research interest. For further information or specific queries about the project, you may contact the supervisor, Dr R Szalai.