|
Project Proposals for William Flynn Scholarship:
IIndex
page
Project Number: 2
Project Title: Self-Organising Intelligent System Generation
Project Supervisor: Dr. Liam Maguire, Dr. Girijesh Prasad, Professor
Martin Mc Ginnity
Background
Research in self-organising systems has been ongoing for the last
50 years particularly in the social science domain and more recently
from a computational science perspective [1]. A self-organised system
is more advantageous than a pre-programmed, deterministic organisation
as the former displays an adaptive character, ensuring that such
a system is reliable and robust. Typically a self-organising system
is comprised of a number of collaborative processing units and such
parallelism decreases the probability of breakdown and ensures redundancy.
Self-organisation is a vital characteristic of any intelligent
system that displays autonomous behaviour. The conventional intelligent
techniques have attempted to mimic certain human characteristics
such as learning (neural networks), reasoning (fuzzy logic) and
finally adaptation/evolution (genetic algorithms). There has been
considerable interest, over the last ten years, in combining these
techniques to produce hybrid approaches which attempt to overcome
some of the drawbacks of the individual techniques [2]. More recently
there has been considerable interest in extending the ability of
these techniques to self-organise particularly in the domain of
fuzzy reasoning and neural networks [3-6]. Such a hybrid approach
is closely related to the developments in adaptive intelligent systems.
Much of the current work in this domain has origins in control theory
where there is a recognition that the control algorithm demands
adaptability characteristics to account for changing environmental
or operating conditions. However, the techniques are readily applicable
to a range of application domains such as medical data processing
and biomedical engineering [7].
The research interests of the Intelligent Systems Engineering Laboratory
(ISEL) have focused primarily on the design, development and implementation
of intelligent computational systems. The Group has recently been
active on projects funded by the European Union (ESPRIT): the promotion
of an open microprocessor user support centre (ECU94k) and a system
level toolchain for embedded performance-critical applications on
a universal microcontroller (ECU398k). More recently the group has
been successful in attracting local industry funding, for example
an intelligent toolkit for the repair and diagnosis of microprocessor
systems (£57k) and two TCS projects. This proposal maintains
this strategy of being involved in innovative and fundamental intelligent
systems research.
Aims and objectives
This project attempts to develop new techniques to generate self-organising
fuzzy and fuzzy neural network systems. The main objectives of the
project can be summarised as:
1. Provide a critical review of the techniques used to develop self-organising
systems.
2. Develop new approaches to how this capability may be implemented
in software.
3. Further develop these novel implementations can be implemented
in hardware and in particular programmable hardware.
4. Design and implement these approaches.
5. Verify and validate the research by implementing a number of
real cases from the biomedical doamin.
6. Improve the design and architecture by review and iteration.
Intellectual Challenge and training and benefits to the student
There are two main intellectual challenges relevant to the project.
The first involves the assessment of the various techniques to generate
self-organising intelligent systems using fuzzy logic and fuzzy
neural networks. The second involves the actual implementation of
the selected strategy and an assessment of its effectiveness on-line
in a dynamically reconfigurable embedded system. There is obviously
a strong training component in the project in terms of the differing
intelligent techniques, assessment and analytical techniques, and
the application domain. The student will benefit from the project
in terms of his/her academic contribution to the emerging disciplines
of intelligent and self-organising systems but also by its application
to an emerging application domain.
Anticipated Outcomes
The successful completion of the project will demonstrate the effectiveness
of the use of self-organising fuzzy and fuzzy neural networks in
a medical application. The project shall develop novel self-organising
intelligent architectures that can be used in this domain. The academic
contribution will be disseminated in a number of academic journals
such as IEEE Fuzzy Systems, Systems Man and Cybernetics and Biomedical
Engineering.
References
1. Yovits, MC et al: "Self-organising systems" McGregor
and Werner, 1962.
2. LP Maguire, TM McGinnity, LJ McDaid: "A Fuzzy Neural Network
for Approximate Fuzzy Reasoning", book chapter in Intelligent
Hybrid Systems: Fuzzy Logic, Neural networks and Genetic Algorithms
by Kluwer Academic Publisher, pp. 35-58, 1997
3. Hoffmann, AG: "On the computational limitations of self-organisation
and adaptation" Neurocomputing, Vol 15, No 1, pp. 69-87, 1997.
4. Rojas I et al: "Self-organising fuzzy system generation
from training examples", IEEE Trans on Fuzzy Systems, Vol.
8, No. 1, pp. 23-36, 2000.
5. Mitra, S; Pal SK: "Fuzzy self-organisation inferencing and
rule generation", IEEE SMC Part A, Vol 26, No.5, pp. 608-620,
1996.
6. Chiang, CK et al: "A self-organising fuzzy logic controller
using Gas with reinforcements", IEEE Trans Fuzzy Systems, Vol.
5, No. 3, pp. 460-467, 1997.
7. Ciresi G, Akay M: "Fuzzy logic in medical control applications"
Biomedical Engineering Applications Basis Communications, Vol.8,
No.6, 25 pp.471-87, 1996.
If you are interested in being considered for a studentship please
contact
the Group Director, Professor T.M. McGinnity by email:
tm.mcginnity@ulst.ac.uk
or telephone: +44-(0)28-71375417.
See the current research section of this website
for details on research projects pursued by existing PhD students
|
