CITS7212: Computational Intelligence

1
CITS7212 Computational Intelligence

• Luigi Barone, Rm 2.12
• Lyndon While, Rm 1.14

2
Unit Overview

• Computational Intelligence (CI) nature-inspired
  heuristic-based algorithms that demonstrate
  emergent adaptive or intelligent behaviour
• Used for solving complex optimisation problems,
  adaptive learning, and knowledge acquisition
• The unit will introduce eight core CI techniques
  
• evolutionary algorithms particle swarm
  optimisation
• neural networks ant colony optimisation
• fuzzy logic artificial immune systems
• Bayesian reasoning learning classifier systems

3
Reading Units

• CITS7212 is a reading unit
• Reading units are less structured than most
  undergraduate units
• The unit coordinators will assist you with
  learning the material, not teach you the material
• The emphasis is on self-learning and
  self-investigation, thus requiring students to
  find more information for themselves than most
  other units
• i.e., the unit will not run as a series of formal
  lectures

4
Unit Structure

• The class will be divided into four teams of 3-4
  students each
• Each team will be assigned two technologies to
  investigate and report on, one of which must be
  used to complete the common project
• different teams will be assigned different
  technologies
• the unit coordinators will advise which of the
  two technologies should be used for the project
• However, all students are expected to gain a
  basic understanding of all eight technologies

5
Unit Structure

• All students will be given 2-3 introductory
  research papers on each technology topic
• The papers will introduce/summarise the topic,
  but will not be sufficient for a complete
  understanding
• Each team is expected to research the topic in
  much more detail than provided by these papers

6
Assessment

• Assessment consists of three components
• three seminars (each worth 10)
• a project implementation and report (40)
• a final exam (30)
• Each team will present three seminars
• an introductory lecture on their allocated
  project technology (weeks 6-7)
• an introductory lecture on their allocated
  non-project technology (weeks 8-9)
• a seminar on 1-2 applications of their allocated
  project technology, with particular reference to
  the project (weeks 10-11)

7
Seminars

• The two technology lectures should introduce how
  the technologies work through a logical
  progression of relevant material
• The applications seminar should provide details
  of applications of the project technology to 1-2
  interesting real-world problems (and the project)
• Seminars will be assessed on the teams
  understanding of the technical material, breadth
  of coverage, and the quality of the instructive
  nature of the presentation
• Each team member must contribute equally
• e.g. one seminar each, or one third of each
  seminar

8
Seminars

• Seminars should run 30-40 minutes
• Aim for about 30-40 slides
• Seminars should not assume any prior knowledge
• Maintain a balance between overview information
  and technical details a good lecture offers
  both
• Ideas should successively build upon each other
• Pictures, diagrams, videos, and demos help
• Additional supporting handouts are not
  compulsory, but may help convey an instructive
  message
• Slides need to be emailed to the unit
  coordinators for inclusion on the unit website
  after the seminar

9
Project Exam

• Project worth 40
• one common project for all teams
• teams must use their nominated project technology
  to create (learn) a solution to the project
  problem
• teams must submit a report detailing the
  applicability of their nominated technology to
  the project problem
• Exam worth 30
• one question from each of the eight CI
  technologies
• students must answer the six questions not from
  their teams technologies

10
Schedule
11
Questions?