A.I./Games Learning Agent.

Design a learning agent which will evolve strategies to survive and compete successfully in Tyrrells Simulated environment. The Simulated Environment was build as part of the PhD research of Toby Tyrrell at the University of Edinburgh in 1993. The intention was to model a complex environment inhabited by a creature with multiple conflicting goals. Tyrrell then implemented several action selection mechanisms, each of which were separately plugged into the creature in order to allow it to choose between the thirty five low level actions available to it. Using the results Tyrrell was able to assess each of the action selection mechanisms and design a new mechanism that out performed all those tested.A new implementation of the Simulated Environment is available as a World-Wide-Mind service. This re-implementation in Java provides precisely the same functionality as was available in Tyrrell's implementation in C.

Intelligent Sheepdog.

The aim of this project is to investigate the simulation of intelligent moving behaviour using sheep and a sheepdog as the example. A program will be developed to simulate the movement of a flock of sheep and a sheepdog. Autonomous Agents will be used to model the sheep and different activities of the sheep, including basic flocking behaviour. The sheepdog will to be modelled as an Autonomous Learning Agent with a Goal to direct the sheep through a specified target point. Testing of algorithms used for the movement and behaviour of the sheepdog and the sheep to see if they are effective and appropriate in doing this. This project should demonstrate emergent behaviour using appropriate rules for the sheep and for the movement of the sheepdog.

Biological Cell Simulator.

This project requires the researching into how a T-Helper cell functions in the immune system. Develop an Agent based model of the dynamic processes involved in T helper cell activation and development. To be able to develop realistic models, and to qualitatively determine the behaviours of complex biological systems, one first needs to recognize the different components/variables experimentally in such systems. Secondly; with enough data/parameter values, more quantitative models can be developed.

Online Quiz Environment

Write a program that will automatically generate Mutiple Choice Computer Architecture tests and handle the submission/marking of results. The Database will include sub-sections for example, Hardware,Number Systems, Logic etc and Users will be supplied with test passwords.

Model of Eutrophication in Lake

Design a system that models and simulates inflow/outflow in a lake. The lake system will include algae which will grow depending on envoironmental conditions (e.g. light, temperature, oxygen, nitrogen). The system will simulate conditions for an Algal Bloom which can deoxygenate the lake water and lead to fish kills in the lake.

Mammographic image processing using wavelet processing techniques

This project involves processing of images using wavelets to enhance the images and further improve detection of cancer. This involves a study on classification of digital mammographic images, using wavelet transform analysis in conjunction with statistical pattern recognition techniques. A comparative evaluation between different wavelet analysis architectures in terms of their classification ability, as well as between different classifiers is to be carried out. The project aims to show if statistical measures clearly distinguish between the normal/abnormal classes, when applied to a large dataset.