Accessibility statement

Neural Networks & Neural Computing - ELE00115M

« Back to module search

  • Department: Electronic Engineering
  • Module co-ordinator: Dr. David Halliday
  • Credit value: 10 credits
  • Credit level: M
  • Academic year of delivery: 2018-19

Module summary

Inspiration and motivation from neuroscience. Structure and function of the human brain. McCulloch & Pitts (MCP) neural computing unit and examples. Perceptron networks, Simple perceptron networks. Computation in the MCP unit: vector product, linear algebra, concept of similarity. Simple perceptron design methods: method of constraints, geometrical design methods. Learning in simple perceptron networks: Perceptron Learning rule. Gradient descent learning and the Widrow-Hoff delta rule. Examples of using formal design methods and WH-delta rule. Associative networks: Structure and function. Associative network dynamics and use as content addressable memories. Associative network storage capacity. Hopfield energy function for associative networks. Hebb's law and unsupervised learning. Unsupervised Hebbian learning, Oja’s learning rule, Sanger’s learning rule and Principal Component Analysis (PCA). Competitive neural networks - Competitive learning rule, Vector Quantization. Self organising feature maps and Kohonen learning rule. Examples and applications of unsupervised learning: familiarity, clustering, feature mapping. Multilayer perceptron networks: structure and function. Formal design methods: tiling. Learning rules for multilayer perceptrons. Backpropagation learning rule: derivation and operation. Example applications of backpropagation. Deep neural networks and Deep Learning, example applications of deep networks. Stochastic neural networks, stochastic units and Boltzmann machines, recurrent network structures and example applications. Spiking neural networks, principles, spiking unit dynamics, example applications.

Module will run

Occurrence Teaching cycle
A Spring Term 2018-19

Module aims

Subject content aims:

  • To provide an introduction to the field of Neural Networks and Neural Computing.

Graduate skills aims:

  • To appreciate the different types of neural network technologies and application areas.

Module learning outcomes

Subject content learning outcomes

After successful completion of this module, students will:

  • Demonstrate an understanding of the structure and function of a neural computing unit.

  • Demonstrate an understanding of the following types of artificial neural networks: Perceptron, Multilayer perceptron, Deep, Associative, Hebbian, Competitive and Boltzmann networks.

  • Demonstrate an understanding of the range of learning rules applied to these networks, including: Widrow-Hoff, Back-propagation, Hebbian, Oja, Sanger, Competitive, Kohonen and Boltzmann rules.

  • Demonstrate an understanding of Spiking Neural Networks.

  • Calculate and use the Widrow-Hoff delta rule for Perceptron networks and the Hopfield energy function for associative networks.

Graduate skills learning outcomes

After successful completion of this module, students will:

  • Communicate understanding and design choices through written content.

  • Use MATLAB to implement and apply a range of neural network algorithms.

  • Apply theoretical knowledge of neural networks to solve problems.

Assessment

Task Length % of module mark
University - closed examination
Neural Networks & Neural Computing
2 hours 100

Special assessment rules

None

Reassessment

Task Length % of module mark
University - closed examination
Neural Networks & Neural Computing
2 hours 100

Module feedback

Students will receive indicative marks and feedback on the examination performance within six weeks.

Indicative reading

+ Churchland, PS, & Sejnowski, TJ, 'The Computational Brain', MIT Press, 1992. ISBN 0-262-53120-8. JBM Library: SK 70 HER, 5 Copies.

++ Hertz, J, Krogh, A, & Palmer, RG, 'Introduction to the Theory of Neural Computation', Addison Wesley, 1991 ISBN 0-201-50395-6. JBM Library: B 3.11 CHU, 6 Copies.



The information on this page is indicative of the module that is currently on offer. The University is constantly exploring ways to enhance and improve its degree programmes and therefore reserves the right to make variations to the content and method of delivery of modules, and to discontinue modules, if such action is reasonably considered to be necessary by the University. Where appropriate, the University will notify and consult with affected students in advance about any changes that are required in line with the University's policy on the Approval of Modifications to Existing Taught Programmes of Study.