saturnmind
GitHub - perseus784/SaturnMind: A repo to understand Q Learning and Deep Q learning
Run grid_environment.py to run just the Qlearning. The experimental setup is simple, it's a grid with 8x8 nodes and some blocks. The aim is to navigate to the diagonally opposite side of the grid with minimal number of steps. As the state space grows and number of actions gets complex, we cannot use the Q table techinque. It will cause a memory overflow or even if you have huge memory, it is just not efficient.