This session is facilitated by Dan Mather, Sam Stuart
About this session
** Note ** This session takes place in the Prototyping Space on Level 9 ** Note **
In this session, you will create your own maze level for an AI to solve through cycles. We will then cut the result out with a lasercutter onto plywood, for you to take home!
The AI will start with no previous knowledge of what the maze is and the outcome requires. It will use algorithms similar to Q-Learning to solve the maze. The AI will solve the maze and will try and find the most efficient route by applying weighting’s to each square and in the end the maze and the most efficient route and level layout will be etched onto a plywood board using a laser cutter.
Goals of this session
For attendees to learn how machines learn through a recursive cycle of trial and error to find the most efficient method of solving a problem through a simple game,
For the attendees to create their own game level that will be tested for the most efficient route from the player’s starting point to the end goal,
For the attendees to be able to take home a laser cut board of their level with the most efficient path etched onto the board.