GridWorld-MDP

The agent lives in a grid. Our agent must go from the starting cell (green square) to the goal cell (blue cell) but there are some obstacles (red squares) blocking the agent’s path.

• Markov Decision Processes
  • Introduction
  • Markov Chain
  • Markov Decision Process (MDP)
  • Rewards and returns
  • Episodic VS Continuous Tasks
  • Policy Function
  • State Value Function
  • State-action Value function (Q function)
  • Bellman equation
  • Value Iteration
  • Policy Iteration
• Installation
  • Prerequisites
  • Clone GridWorld’s repository
  • Install requirements
  • Docker
• Usage
  • Play game
  • Docker
• API Reference
  • gridworld package