Aims & Learning Objectives:
Aims: To provide the fundamental principles of various artificial intelligent techniques and insights of how to apply these techniques to solve practical problems. In particular, the course provides in depth knowledge of one of the most popular artificial intelligent technique - neural network, with detailed practical implementation procedure and extensive application examples.
Objectives:
After completing this module, students should be able to: distinguish the differences between intelligent techniques and conventional techniques; be aware of the opportunities where intelligent techniques might be most beneficial; be able to construct simple intelligent systems to solve practical problems; be able to further enhance the performances of intelligent techniques.
Content:
Neural Networks (NNS): artificial neurons and neural networks; learning process: Error-correction learning, Hebbian learning, Boltzmann learning, competitive learning, supervised/unsupervised learning; Perception and multilayer perception; self-organising Kohonen networks, Kohonen feature maps; Hopfield neural networks; practical implementation and applications: the electronic nose, fault diagnosis/classification in engineering networks. Expert Systems (ES): major characteristics of expert systems; knowledge representation techniques; inference techniques; rule-based expert systems; applications in power systems. Fuzzy Logic (FL): fuzzy set theory; fuzzy inference; fuzzy logic system; fuzzy control; applications on power systems. Genetic Algorithms (GA): adaptation and evolution; genetic operators; a simple genetic algorithm; genetic algorithms in optimisation and learning.
| |
Fax: +44 (0) 1225 386366