Solving the Rubik's Cube with Learned Guidance Functions

This paper introduces move sequence problems-problems where a system can exist in a number of states, including a goal slate, with moves between those slates. This paper introduces Learned Guidance Functions (LGFs) as a machine learning method to tackle these. An LGF is a function learned by supervised machine learning that predicts how far a particular state is from the goal state. These methods are applied to the challenging problem of unscrambling a Rubik's Cube.