The group contains academics from the University of Aberdeen’s knowledge technologies group, the University of Dundee’s computational sciences and interactive digital media groups and RGU’s knowledge-based systems and computational intelligence groups. These researchers cover the topics of knowledge, computation and digital media that are central to Intelligent Systems. Modern society makes vast quantities of information available to decision-makers. The development of reliable and ubiquitous computing makes it possible to collect data and events across large systems and organisations, such as the health service, transport systems and multinational corporations. Monitoring equipment and microcontroller technology provide the capacity for control of complex industrial processes. Technological advances in genetic analysis such as micro-arrays are making available vast biological data sets for analysis and interpretation. The Web has made the possibility of reusing existing knowledge sources in a different context, a tantalizing prospect.
All of these areas give rise to difficult decision-making problems posing problems of immense complexity. Typical features of such problems are complex interaction of constraints, combinatorial explosion in solution space, multiple and conflicting objectives, non-deterministic outcomes, the use of predictive simulations to evaluate outcome, incomplete or poorly structured knowledge, noisy or missing data, and operation in a distributed environment.
Intelligent systems have long been applied to such problems. Persistent themes of this research are: knowledge representation, acquisition and refinement, knowledge discovery, experimental design, grammar acquisition, statistical modelling, parameter optimisation, comparison of intelligent algorithms and the appropriate selection for particular problems, self-adapting systems, integration between algorithms and human decision-making in mixed-initiative systems, and frameworks for applying techniques to new problems.
The Intelligent Systems grouping brings together, from three institutions, researchers with a well-established international reputation in Knowledge Technologies: discovery and representation, capture, refinement, re-use and transformation. Our significant activities in this area include data and text mining, image understanding, knowledge bases and constraints, ontological modelling and their transformation and subsequent use in intelligent systems. The group has a strong international reputation in Machine Learning encompassing: self-adapting and biologically-inspired systems, grammar acquisition, experimental design and parameter optimisation. Our significant activities in this area include evolutionary algorithms, inductive logic programming, Bayesian probabilistic models, graph and combinatorial algorithms, and particle swarm optimisation and their use to build intelligent systems.
The grouping has a wealth of experience in developing intelligent systems for tasks that have highly-demanding requirements including multi-objective optimisation, non-linear constraints, learning and inference from image data, complex knowledge representations, computational complexity and distributed knowledge bases. Our expertise has been refined on a wide range of real-world application areas including medicine and healthcare, bioinformatics, pharmaceuticals, planning and scheduling, image analysis, surveillance and engineering.
Patrik Holt, RGU
Manuel Trucco, UoD
Tim Norman, UoA