Artificial Intelligence (AI) is the design and use of computer systems to understand and generate human-level decision making. These systems represent, reason with, and learn from, different descriptions of knowledge and uncertainty. It is also the design and use of such computer systems, and their associated algorithms and architectures, to solve practical problems in different domains. 
Our group hosts fundamental research into theoretical properties of AI algorithms and AI problems, aimed at advancing the scientific understanding of AI. We also lead methodological research exploring how our insights and algorithmic innovations can drive progress in different scientific disciplines.
Machine Learning

Generalisation analysis and uncertainty estimation
Understanding the theoretical underpinning of learning methods, development and analysis of algorithms that enable valid conclusions from empirical data.
Evolutionary & Nature-inspired Computation

Runtime and dynamical analysis and multi-objective optimisation
Development and analysis of methods inspired by natural systems such as evolution, ant foraging, molecular computation and cell signalling.
Computational Modelling

Neural Networks and Complex Systems
Analysis of properties of complex models, development of domain-specific models in healthcare, natural language, bio-medicine, industrial automation, cyber security, finance, astronomy and environmental science.
Data, Streams and Pattern Mining

Extracting meaningful patterns, insights and relationships in large datasets
Data mining, data streams, time series, and pattern mining study algorithms aimed at extracting meaningful patterns, trends, relationships, insights, and knowledge from possibly large, complex, or time-evolving data sets.
Planning and Multi-agent Systems

Complex decision-making and coordination challenges in dynamic and interactive environments
Planning, Multi-agent Systems, and Game theory provide the building blocks for addressing complex decision-making and coordination challenges in dynamic, and interactive environments.
Members of our group draw methods and inspiration from a wide range of fields, including statistics, high-dimensional probability theory, random matrix theory, stochastic processes, information geometry, dynamical systems, theoretical computer science, game theory, evolutionary biology, and the physical, chemical, and brain sciences. As AI continues to transform the world, we leverage these diverse perspectives to establish rigorous foundations in its core areas, such as:
To learn more about the exciting work we do, please visit our members’ homepages.
The group has a mailing list. If you would like to join the group, please email
Peter Tino or
Ata Kaban outlining your interests.