| Popis: |
Rubik’s Cube is a classic and standard puzzle which has a very large state space of 4.3 × 1019 different states in 3 × 3 × 3 cube. However, there is only one terminal state. Recently, there are few contemporary solutions to solve Rubik cube which exploit Machine Learning Techniques. Our goal is to explore and generate Reinforcement Learning, CNN and LSTM techniques for sequence learning of Rubik cube solution with entropy modelling. The entropy is maximum when the cube is in unsolved state and minimum when it’s in solved state. We perform an Entropy traversal modelling for Rubik cube solution. |
