1 x 2-hour lecture weekly
1 x 1-hour seminar weekly
1 x 2-hour computer lab weekly
4 x 13-hour project works per semester
1 Admission into MIT-Master of Information Technology
2 COMP8702 - Computer Programming 1 GE
3 Admission into MSCCS-Master of Science (Computer Science)
3a Admission into GDPCSC-Graduate Diploma in Computer Science
3b Admission into MCS-Master of Computer Science
3c Admission into MCSAI-Master of Computer Science (Artificial Intelligence)
3d Admission into MDSC-Master of Data Science
4 1 of COMP1102, ENGR1721, ENGR8800
Must Satisfy: ((1 and 2) or ((3 or 3a or 3b or 3c or 3d)) or (4))
Enrolment not permitted
1 of COMP2712, COMP7715 has been successfully completed
Assumed knowledge
Basic knowledge of computing such as found in COMP8701 Fundamentals of Computing GE and basic knowledge of programming in an imperative and object-oriented language such as Java and C/C++.
Assignments, Exercises
Topic description
This topic provides a thorough exploration of the heuristic optimization techniques of computational intelligence, with a working language of C/C++ for examples and assignments.

Specific areas covered include:

  1. Theory and practice of neural networks (NN) including feedforward networks, backpropagation, learning machines, self-organizing networks, and deep learning

  2. Theory and practice of evolutionary algorithms (EA) including genetic algorithms, genetic programming, pareto optimization and selection methods

  3. Theory and practice of fuzzy methods (FZ) including fuzzy logic, type 1 and 2 models, and soft clustering

  4. Theory and practice of Bayesian methods (BY) including Bayesian statistics and belief networks (BN)

  5. Practical experience writing, deploying and evaluating heuristic optimization algorithms

  6. Practical applications of heuristic, reasoning and optimization techniques in areas such as speech and language processing, character and image recognition, data base and data mining

  7. Practical understanding of the application nature-inspired methods and the process of modelling and simulating natural processes both for general computing use and in the natural domain.
Educational aims
This topic provides a theoretical and practical introduction to the development and application of heuristic and nature-inspired techniques common in Computational Intelligence (CI). Students will understand the mathematical and algorithmic principles behind CI and gain hands-on experience in employing CI techniques towards solving practical problems in a variety of domains. The scope of the topic incorporates increasingly relevant applications such as biometrics, speech and computer vision and the ethical issues associated with these. Students will also obtain skills in analytical thinking, experimental research, and effective communication within the wider context of CI.
Expected learning outcomes
On completion of this topic, students will be expected to be able to:

  1. Understand literature and applications associated with Computational Intelligence (CI) and communicate effectively in this field

  2. Appreciate the common principles and best practices linking the constituent domains of CI

  3. Utilize and integrate CI techniques towards solving a variety of relevant real-world problems

  4. Evaluate and compare in a scientific manner the suitability and effectiveness of different CI techniques

  5. Deal with the practical and ethical constraints associated with CI-based systems.