PhD Studentship - Human-AI Collaborative Representation Learning

The University of Manchester - Computer Science

Qualification Type: PhD

Location: Manchester

Funding for: UK Students

Funding amount: £20,780

Hours: Full Time

Placed On: 25th June 2025

Closes: 15th August 2025

This 4-year PhD project will be funded by DLA studentship. UK students and EU students with settled status are eligible to apply. The successful candidate will receive a tax free stipend set at the UKRI rate (£20,780 for 2025/26) and tuition fees will be paid. The start date is 1st October 2025.

Visualising data objects, by embedding them into a 2 or 3-dimensional space through a representation learning algorithm, has been widely used for data exploratory analysis. It is particularly popular in areas such as biology, chemistry, psychology and social science, facilitating knowledge discovery. The intuitively uninterpretable high-dimensional data and network data become visually scrutable upon being mapped into 2 or 3-dimensional spaces, enabling insights about the underlying structure and distribution of the data. However, due to the heavy data compression into a space with only 2 or 3 degrees of freedom, information loss is inevitable, thus it is natural to drop “unimportant” data patterns in a visualisation result.

In practice, the desired data pattern and structure to preserve in a visualisation result can vary across domains due to the different nature of the downstream decision-making tasks that the visualization results will serve. An important research question is how to enable a representation learning algorithm to develop an ability to choose what main data pattern/structure to preserve?

This PhD project will approach this question by developing modelling strategies and pipelines to enable human-in-the-loop representation learning algorithm design, asking users to decide what data pattern/structure should be preserved by an algorithm and injecting the preference into algorithm design. The visualisation results will be tested by examining how well it can assist a series of decision making tasks through collaborating with experts from the problem domain.

Applicants should have, or expect to achieve, at least a 2.1 honours degree or a master’s (or international equivalent) in a relevant science or engineering related discipline. Strong background/skills on machine learning, mathematics, probabilistic modelling and optimisation are preferred.

To apply, please contact the supervisor, Dr Mu at Tingting.Mu@manchester.ac.uk. Please include details of your current level of study, academic background and any relevant experience and include a paragraph about your motivation to study this PhD project.

How to apply: uom.link/pgr-apply-2425