Learning of a maze using a genetic algorithm

Abstract

We attempt to realize the ability of a robot to acquire the intention of a maze using a genetic algorithm and decision making. In this case the intention means the rule peculiar to a maze. We made a robot learn in a maze based on an intention, and made it run in two mazes, one based on the same intention and the other based on a different intention. The results obtained confirm that learning of an intention has been achieved.<>