The group has a broad interest across artificial intelligence, machine learning and data science. Our extensive research expertise includes a number of topics in applied and theoretical research.
Contact: Dominique Chu
Members of the group routinely work with industry to solve practical problems, for example via funded KTP projects or consulting. If you would like to work with us, you can contact a relevant member of staff directly, or alternatively, visit the University Business and Innovation Gateway. We also welcome enquiries from students who are interested in doing a PhD with us. Our expertise includes the following topics:
We develop algorithms to enable machines to automatically recognise patterns in data in supervised and unsupervised settings. We are also interested in reinforcement learning.
Explainable and Trustworthy AI
State of the art machine learning algorithms are most commonly black-box models. This means that we cannot know why an algorithm has come to a particular decision. This prevents their application in domains where it is important to justify how decisions have been made, as for example in finance or medicine. Our group researches novel algorithms that perform well and are explainable.
Computational neuroscience / brain-computer interfaces
We use machine learning and advanced statistical methods to analyse physiological signals (such as EEG data) to understand the brain. These can then be used to understand how the brain works, but we also build systems that allows computers to be controlled mere by thoughts.
We find novel methods to automatically visualise complex information, use graphs to represent semantic information and use graph theory to better find hidden patterns in data.
AI can be used to create new, interesting artefacts and art. It is not clear, however, what it means for a computer to be creative and how human creativity differs from machines that produce art. Our research addresses these questions, with particular interests in music informatics, language-based creativity and creativity theory.
Smart environments and IoT
The proliferation of smart sensing and IoT technologies opens new opportunities and challenges for the development of novel intelligent systems with real-world applications. Working with real-world data, there is a need for the development of novel data analytics and machine learning techniques that can process large-scale datasets form sensors and smart devices and infer knowledge about systems performance or human behaviour. Applications can range from data analytics for healthcare, smart cities, and smart buildings, to interactive museums and performing arts.
Applications of machine learning and data science to healthcare
Data science and AI can be used to find better treatments for diseases, new drugs or improved care. We have a strong portfolio that ranges from machine aided diagnostic, assistive living and smart environments, to the use of virtual reality to improve the lives of Alzheimer patients. We have expertise developing Computer Aided Detection systems for radiological images and in biomedical data classification. We have a long track record of working closely with health specialists to explore novel solutions that can have significant impact on the quality of care offered to patients.
In our Cognitive Robotics and Autonomous Systems (CoRAS) lab, we research how machines can be programmed to work collaboratively and productively with one another and alongside humans. Our research in cognitive and developmental robotics is aimed at designing intelligent behaviour in robots, including high-level perception, motor, language and cognitive skills. We are working towards social and assistive robotics, mobile robotics and human-robot interaction.
Natural language processing and understanding
We have interests in language processing, understanding and generation through a breadth of techniques based on probabilistic modelling and neural networks. We study state of the art transformers, language models and cognitive architectures inspired by cognitive science, linguistics and psychology. Our research covers mono (English language) and multilingual approaches.