davidjegan/Game-play-using-Reinforced-Learning

This code addresses how we could teach an agent to navigate in a grid-world environment. In this modelled tom and jerry game, we apply reinforcement learning DQN (Deep Q-Network) to make the agent find the optimal shortest path from the goal(Jerry) to initial position(Tom) from its history of interaction with the environment. These two initial positions are deterministic.

Jupyter Notebook