Visualising Patterns in Changing Information
The aim of this project is to develop and evaluate new techniques for visualising dynamic data (e.g. stock market). As part of this project a generic 3D visualisation tool will be developed and applied to data from a range of application domains (stock market, sport, neurophysiology, consumer demand, etc).
An Agent-Based Model of Emotion
The aim of this project is to develop a model of conflicting emotion which can be integrated into non-playing computer game characters. It examines current models of emotion using traditional conceptual modeling tools derived from software engineering. The model will be integrated and tested using agent-based simulations.
Navigating Knowledge using a Metro Map Layout
Metro maps were first introduced on the London Underground and now have become a common place tool for navigating transport routes around the world. The question explored in this project is; can more abstract knowledge be represented in this way? For example, would such a layout help novice users navigate through more abstract content. For example, the content of an instruction manual.
A Game Theoretic Model of Renewable Energy Generation
This project looks at a possible future for Australia when many small renewable energy generators are installed on the electricity grid and partake in a self-interested market. In this scenario each generator tries to optimize its cost recovery in the face of real time changes in electricity price and a regulatory environment designed to encourage renewable energy generators to support the grid during times of high demand or faults. The project will draw on multi agent theory, game theory and optimisation techniques.
Asynchronous Boolan Networks as Models of Gene Networks - What Can They Tell Us?
This project looks at boolean models of gene networks, where each gene is modeled as either being on or off. These models have already provided new insights into the stability of critical functions such as cell division. Once the assumption of synchrony is relaxed the models have the potential to provide more information about stability. This project will build on network theory and make use of multi agents models.
New Measure of Heart Rate Variability for Disease Detection
A person's heart rate can be measured using simple and cheap sensors, but this only provides the time between each heart beat, not a full electrocardiograph. However, it is becoming increasing apparent that measures calculated from the heart rate can be very useful for the diagnosis of a number of diseases, so could be used as simple health screening system. This project will investigate new measures derived from heart rate recordings and will use some ideas from artificial intelligence.
Automated Feature Selection for Data Analytics in Large Databases
In large databases, there may be many fields that could be used in analysis such as cluster and outlier detection and association analysis. Including all the fields assumed to be relevant has been shown to be sub-optimal and can actually result in poor performance. However, it is difficult to know which fields to select. This project uses search and optimisation methods to determine the optimum set for a database, given the query type.