Research projects in Information Technology

Overview

This project proposes a novel quantum-enhanced learning analytics framework for higher education, focusing on early identification of at-risk students and optimisation of intervention strategies using hybrid quantum-classical approaches. While current learning analytics systems rely on classical statistical and machine learning techniques, they often struggle to capture the complex, uncertain, and multi-dimensional nature of student learning behaviours.

Background and Motivation

Modern deep learning models have achieved remarkable success in computer vision and natural language processing. However, they typically produce overconfident predictions and lack reliable mechanisms to quantify uncertainty. This limitation becomes particularly problematic in high-stakes applications, such as healthcare diagnosis, autonomous systems, and scientific discovery.

Project Background & Motivation

Project Background and Motivation

The "inverse design" of molecules from analytical spectra (such as MS2, NMR, or IR) is a fundamental bottleneck in analytical chemistry, metabolomics, and drug discovery. While deep generative models have shown promise in proposing novel molecular structures, they typically require massive, cleanly labelled datasets to train effectively.

This PhD project focuses on the design and evaluation of hybrid quantum–classical algorithms for large-scale data analytics and optimisation problems.

The research will investigate how quantum computational techniques can be combined with classical systems to improve performance, scalability, and solution quality for tasks such as:

We are seeking a motivated PhD candidate to work on unsupervised music emotion tagging within the broader field of affective computing. The project aims to develop reproducible machine learning approaches for automatic emotion recognition in music, with stronger theoretical grounding, transparent model implementation, and rigorous validation.

Using the Project-1 hiPSC platform, this project builds AI pipelines to learn disease-relevant representations from cellular images, fused with multi-omics. Models will classify diagnosis and predict treatment response with strict donor-level splits, cross-regional external validation, and fairness audits (sex/ethnicity stratification). Interpretable AI (e.g., attribution maps, SHAP) will nominate mechanism-anchored biomarkers and candidate interventions. Tooling will be containerised and open to support reproducibility and clinical translation.

Digital health (DH) technologies promise significant improvements in healthcare delivery, but their successful implementation requires rigorous evaluation to ensure they meet clinical and operational needs.

There is a key opportunity for specialist work in an emergent intersection area which we can call Value-Based Digital Health (VBDH).

To expand further, VBDH is a discipline area that sits at the intersection of DH, HTA and IS and is focused on defining, delivering and measuring value in all parts of the lifecycle of DH products. (Prof C Bain, 2024)

In order to achieve ubiquitous success in AI for Pathology around the world (in particular in histopathology), a number of key technical issues and challenges need to be addressed. These - and potential remedial approaches - will be explored in this research program and they include (as examples): 

This research focuses on developing and evaluating methodologies for the optimal design of control charts within the framework of Statistical Process Control (SPC). The study aims to determine the best configuration of chart parameters, such as sample size, sampling interval, and control limits, to minimize detection time for process shifts while controlling false alarm rates. It explores both traditional and advanced optimization techniques, including analytical models, simulation-based approaches, and data-driven algorithms.