[jobs] PhD studentship: Intrinsically-aligned machine learning

[Fabio Cuzzolin]

Fully funded Leverhulme PhD studentship: Intrinsically-aligned machine learning

Oxford Brookes Institute for Artificial Intelligence, Data Analysis and Systems (AIDAS), Oxford Brookes University

Deadline: October 31 2025
Duration: 3 years, international fees covered
Bursary: £19,237
Start date: January 2026

Project Description: Whereas traditional machine learning is solely interested in model selection (i.e., identifying, given the available data for the task at hand, the model that is expected to perform best), we propose a new paradigm for an "intrinsically-aligned" artificial intelligence, where accuracy, fairness and explainability are all taken into account when selecting the "best" AI model.

In a truly cross-disciplinary effort, this project, funded by the Leverhulme Trust and in collaboration with Carlo Ludovico Cordasco from The University of Manchester, will leverage results from human decision-making to inform the design of this new paradigm, and feed the results of the latter back into human decision-making to help make it more explainable.

The PhD student will: (1) develop novel performance metrics combining accuracy and explainability, to be tested across different AI model types; (2) devise new algorithms for selecting models optimised for holistic performance, combining both accuracy and explainability; (3) extend statistical learning theory to offer theoretical bounds for intrinsically-aligned AI models; (4) employ the newly-developed metrics to train deep neural networks which are intrinsically explainable; (5) design a new multi-dataset benchmark for assessing the trade-off between accuracy and explainability.

Essential selection criteria:

·      At least an upper second class degree (preferably MSc) in a Science or Technology discipline.
·      Good working knowledge of machine learning and deep learning.
·      Hands-on knowledge of Python or PyTorch for implementing machine learning and/or deep learning algorithms.
·      Capability to work both independently and as part of a team.
·      Excellent written and oral communication and organisational skills. Proficiency in written English is required.
·      A real passion and commitment for research.

Desirable criteria:

·      Knowledge of a variety of deep learning architectures and methods.
·      Knowledge or past work on explainability in AI.
·      Previous publication record in relevant fields: AI, machine learning, computer vision, etc.
·      Previous successful project on a relevant topic.
·      Good knowledge of statistics, probability or statistical learning.

Contact: Prof Fabio Cuzzolin, fabio.c...@brookes.ac.uk

To apply, please email Prof Fabio Cuzzolin and send: (1) your up-to-date CV and (2) a brief statement of research interests, describing how past experience and future plans fit with the advertised position and the project.

Interview dates: November 3-7