The Advanced Programming unit in the ACT Senior Secondary Curriculum provides students with an opportunity to encounter computer science concepts usually reserved for university education. At Gungahlin College, the unit has focused on providing students with a spattering of these concepts (such as recursion, regex and big-o notation), and then allows the students to identify a field of computer science to dive into. This is spaced out over a semester, requiring the students to research their chosen computer science concept and develop their own application capable of providing a practical implementation of that concept.
Two examples of this are Justin Cragg’s project exploring Neural Networks and their application to AI in games, and Dexter Saemo & Alen Zejnic, and their investigation into Pathfinding, examining the Dijkstra and A* algorithms.
Learning AIs, specifically those which use Neural Networking, are an effective way to mirror human thought processes. This allows for a wide range of simulations which rely on varied input to create realistic reactions to specific situations. Justin’s project demonstrates how a learning AI can be used to mirror a simple organism’s actions and reactions towards various environmental stimulus. In the simulation a simple creature has to determine the best way to survive in an unknown environment, while considering both short and long term consequences for its choices. The simulation can allow for both biological and psychological analysis, through the studying of the creatures’ behavioural and learning patterns.
Dexter and Alen explore both Dijkstra’s and the A* path finding algorithm through development of a pygame based implementation of a simple environmental exploration. The application of these algorithms underlies many large scale AI technologies currently in use in the world, and form the basis of their discussion on real world and gaming applications of artificial intelligence.