Mathematical Institute, University of Oxford Oxford Postdoctoral Research Associate 8 December 2025

Machine Learning and Data Science

• Learning with Structured & Geometric Models - apply tools from manifold learning and Riemannian optimisation to better training and novel network designs.
• Low Effective-dimensional Learning Models - extend foundational theory to reduce trainable parameters without sacrificing accuracy, analysing low dimensional building blocks.
• Implicit Regularization - develop mathematical understanding of implicit regularisation in deep neural networks to guide algorithmic paradigms that combine statistical optimality and computational efficiency.
• Reinforcement Learning through Stochastic Control - develop stochastic control methods providing mathematically grounded continuous time RL, addressing scalability for high frequency observations.

This is a two year fixed term position funded by a research grant from the EPSRC. The start date is flexible.

• Conduct research within the remit of the large scale project, collaborating with other hub members at Oxford and partner institutions.
• Publish results with co authors in refereed journals and proceedings.
• Contribute a small amount of teaching (at most three hours a week during academic terms).

Applicants must provide two referees who will send reference letters directly to before the closing date.