Intelligent Systems have attained a permanent and secure role in industry. Since their earlier successes in 1980, many systems have been developed and deployed in industrial and commercial sectors. A good understanding of how these systems are developed, exposing their strengths and weaknesses, is presented through a balanced mixture of lectures, practicals and discussion. Techniques presented thus add to the repertoire of problem-solving strategies for the organisation.
This course will give the participants an introduction to a range of techniques and skills that are required to develop intelligent systems. All the theoretical concepts are supported by a series of practical sessions.
The course is intended for practising scientists and engineers who wish to exploit the benefits offered by Intelligent Systems. No specialist knowledge of the area is assumed. Participants should normally have a degree in a scientific discipline with good working knowledge in a higher-level programming language.
The course includes:
Overview of Intelligent Systems:
A survey of intelligent
systems used in industry and research, characteristic features of intelligent
systems, discussion of software tools for the development of
intelligent systems.
Survey of problem- solving approaches.
Knowledge Representation:
Discussion of types of
knowledge and knowledge representation techniques - object - attribute - value
triplets, rules, semantic networks, frames and first order logic. A case study
in blackboard systems.
Inference Techniques:
Methods of reasoning by hurmans
- deductive, inductive, abductive, anaolgical, common-sense and non-monotonic
reasoning. Application of Dempster-Shafer Theory for approximate
reasoning.
Bayesian Inference:
Elementary introduction
to Bayesian Theory and Bayesian Belief networks.
Fuzzy Logic and Fuzzy Inference:
Introduction to fuzzy
logic and fuzzy inference techniques. Development of simple fuzzy intelligent
systems.
Knowledge Acquisition:
Sources of knowledge,
knowledge elicitation tasks. Common difficulties in acquiring knowledge -
psychological view of elicitation problems, problems with relevant, incomplete,
incorrect and inconsistent knowledge. Interviewing techniques and analysis of
knowledge. Practical session on knowledge acquisition.
Software Tools and Languages:
Tutorial introduction
to CLIPS (C Language Integrated Production System). Development of a small
"expert" system in CLIPS including knowledge acquisition. Introduction to SOAR
(a unified architecture for developing intelligent systems). An overview of AI
languages. Practicals on Fuzzy Logic are based on MATLAB Fuzzy Logic
Toolbox.
Practicals:
All the major topics will
be accompanied by tutorial sessions. No prior knowledge of these languages is
assumed. However, working knowledge in a higher level language and familiarity
with UNIX workstations is desirable.
The course lectures will be given by the teaching and research staff of the Applied Mathematics & Operational Research Group under the direction of Dr Venkat V S S Sastry with the assistance of other colleagues. The programme normally includes a distinguished speaker from industry or academia.