Tagged With reinforcement learning

Meet DeepCube, an artificially intelligent system that's as good at playing the Rubik's Cube as the best human master solvers. Incredibly, the system learned to dominate the classic 3D puzzle in just 44 hours and without any human intervention.