Authorised by Academic Registrar, April 1996
Objectives At the completion of this subject students should: know basic neural network architectures, their operation and applications; understand the tasks and techniques involved in development of neural network applications; be able to apply neural network environments to develop neural network applications; and appreciate how neural computing fits in with traditional computing paradigm.
Synopsis This subject provides students with a broad understanding of neural computing, also known as neural networks. There will be a strong emphasis on application of neural networks in business and industry. In particular the following topics will be covered: the history, architecture and biological motivations; learning in neural networks, the perceptron and its limitations, backpropagation, Kohonen self-organising map, counterpropagation, Hopfield network, bidirectional associative memories (BAM) network, fuzzy logic, neuro-fuzzy systems, neural networks implementations, applications of neural networks; neural network expert systems, neural networks as decision support tools, engineering applications, development of neural network applications using interactive environments, data preprocessing, building heuristic and techniques, training, testing, evaluating neural network performance.
Assessment Practical assignments: 60% + Unit test: 40%