This is an existing unit, which has been reviewed and revised. The content of this unit has been changed in order to give emphasis on real-world business and economic applications and to provide the sound theoretical and practical knowledge of recent neural network and evolutionary computing techniques.

This unit will replace existing unit GCO4015

07/12/2017 - Unit disestablished at FEC 5/17 Item 7.3.

This unit will include AI techniques for OCL students

This unit is for OCL students

AI techniques are used in every field to make intelligent decisions, so this unit will cover AI techniques for OCL students

Upon successfully completion of this unit, students will:

1. understand the underlying theories of neural networks and evolutionary computation and their typical applications.
2. be familiar with the basic tools and terminology in the fields.
3. have the skills required to identify the type of problems and subsequently select the most suitable technique.
4. be able to analyse, solve, model, and simulate real-world business and economic problems using the acquired knowledge.
5. have practical experience using the required tools to solve business and economic problems.

ASCED : 020119

Introduction to neural networks and their applications. Simple neural networks for pattern classification. Multilayered neural networks (backprogration and its variations for faster training and adaptive architectures). Unsupervised neural networks (Kohonen Self Organising Maps). Case studies. Introduction to evolutionary computation and its possible applications. Genetic algorithms. Modeling and simulation with genetic algorithm in economic systems. Genetic programming and design issues of evolutionary algorithms. Hands-on experience to solve real-world business and economic problems using available software tools.

Introduction to neural networks and their applications. Simple neural networks for pattern classification. Multilayered neural networks (backprogration and its variations for faster training and adaptive architectures). Unsupervised neural networks (Kohonen Self Organising Maps). Case studies. Introduction to evolutionary computation and its possible applications. Genetic algorithms. Modeling and simulation with genetic algorithm in economic systems. Genetic programming and design issues of evolutionary algorithms. Hands-on experience to solve real-world business and economic problems using available software tools.

Prescribed Text:

Kate A. Smith, Introduction to neural networks and data mining for business applications, Eruditions Publishing, 1999, ISBN 1-86491-004-6.

Recommended Texts:

• Laurence Fausett, Fundamentals of neural networks: architectures, algorithms, and applications, Prentice-Hall, 1994.
• David Fogel, Evolutionary Computation: Toward A New Philosophy Of Machine Intelligence, 2nd ed., IEEE Press, 2000.
• T. Back, D. B. Fogel, and T. Michalewicz (Editors), Evolutionary Computation 1: Basic Algorithms and Operators, Institute of Physics Publishing, Bristol and Philadelphia, 2000.
• Herbert Dawid, Adaptive Learning by Genetic Algorithms: Analytical Results and Applications to Economic Models, 2nd Edition, Springer, 1999, ISBN 3-540-65607-3.

Software Tools:

• NeuNet Pro 2.2 or neural network toolbox of Matlab 6.0.
• The Nenet version 1.1 for Kohonen Self Organizing Maps.
• An optimisation tool based on genetic algorithm (GOAL).

Off campus only

The following study materials will be provided:

• On-line study materials for each week.
• On-line tutorial problems that cover both theoretical and practical aspects in order to reflect the unit objectives and assess the students' knowledge.
• The printed unit guide outlining the subject objectives, requirements, and assessments.

The practical problems for both tutorial and assignments will be selected from real world business and economic applications. Software tools will be needed to solve those problems.

Newsgroups, emails, telephone and fax will be used for the interaction betwen the students and the lecturer. WebFace assignment systems will be used for assignment submission.

Weekly study materials will be provided during the semester and these will cover the fundamentals of neural networks and evolutionary computation (objective 1) as well as objectives 3 and 4. Tutorial streams and assignments will provide the students with the opportunity of assessment of theoretical knowledge, becoming familiar with the available software tools (objective 2), and hands-on experience to solve the business and economic problems using the respective tools (objective 5). They will also reflect the objectives 3 and 4.

Two assignments : 40% -- Examination (two hours): 60%

Assignment 1 will assess the theoretical concepts of neural networks and the capability of solving a real-world practical problem by analysing, modelling, and using neural networks software tools.

Assignment 2 will evaluate the theoretical underpinning of evolutionary computation and the ability of analysing, modelling, simulating, and solving a business or economic problems using genetic algorithm tools.

A two-hour examination will assess the theoretical knowledge of neural networks and evolutionary computation and their typical applications. It will also estimate the students' skill to identify, analyse, and model the real-world business and economic problem.

Students are expected to spend an average of 12 hours per week on this unit.

NA

NA

NA

Approximately 0.5 EAS (Effective Academic Staff) for one semester including marking.

Replacing existing unit GCO4015, so there are no new library requirements

Gippsland School of Computing and Information Technology - 100%

NA

None

None

None

None