OKeefe_old

Simon O'Keefe

Introduction

My research interests are primarily in the area of neural network architectures and the application of neural networks. I have a long standing interest in binary neural networks in particular. These are networks of simple processing elements (nodes) in which the links between nodes have a weight of either 0 or 1, unlike "traditional" neural networks that use continuous weights.

Recent work I have been involved in has been the application of neural networks to data mining of various forms. Most recently, I have looked at the use of binary neural networks as a basis for identification of unusual cases in large customer databases, possibly indicating fraud.

Other areas of current interest are:

1. Aspects of bioinformatics, particularly in relation to the use of neural networks in bioinformatics, but also the use of binary networks for cell modelling;
   
2. Cellular automata and agent-based models of biological systems
3. Visualisation of, and user interfaces to quantum computing systems.
   

Use of neural networks in data mining.

The use of binary neural networks for representing classification structures and the limits on the accuracy which can be achieved under these conditions: The types of classification structures that might be implemented are decision tables or decision trees. Given an implementation of such a classifier on a binary neural substrate, what are the limits on the accuracy of the system, and how might the usefulness of the system be maximised by combination with other (perhaps slower, more expensive but more accurate) systems?

The use of self-organising networks for data mining

Particularly where the self-organisation is performed by a binary network. Self-organisation as a principle is exciting because it allows the data modeller/miner to minimise the influence of preconceptions on the analysis of the data. However, interpretation of the model is then more complex because of the lack of assumptions about the structure in the data.

Construction of neural network models based on cognitive psychology models.

Neural networks are based on a particular view of how the brain operates at a low level. Success in the application of neural networks has been mixed, despite results showing the computational power of the networks, because of this low-level emphasis in the design of the network architecture. By inclusion of information about the operation of the brain at a cognitive level, we might hope to generate more successful artificial neural models of computation.

Chemical substrates for neural networks

An exciting recent development elsewhere has been the introduction of chemical substrates for computation. recent published work has included some suggestions for the construction of single neurons. I would be interested in supervising projects to improve on the current models.

Bioinformatics

Computational properties of binary networks. Binary (or boolean) networks (not the same as binary neural networks) have been studied elsewhere for some thirty years. They are graph models for representing the interaction between aspects of a system. For instance, the interaction between genes in a cell may be represented by a graph indicating which genes' activity influences which other genes. In this model the activities are binary - on or off. The overall behaviour of the system is complex, and quite often may only be determined by simulation. There are many results relating to the number of attractors (cyclic states) such a system will settle into, the sensitivity to disturbance and so on. What I am interested in is how useful such systems are as computational or control devices.