This repository contains a Python implementation of Monte Carlo Policy Evaluation applied to a simple gridworld environment.

The Gridworld class represents a 5x5 grid environment with special states ('A' and 'B'), each having specific rewards and transition dynamics. The agent can take actions ('north', 'south', 'east', 'west') to navigate through the grid.

The MonteCarloPolicyEvaluation class is designed to evaluate the state values using the Monte Carlo method. It randomly generates episodes using a random policy and updates the state values based on the observed returns.

To run the Monte Carlo Policy Evaluation on the gridworld, follow these steps:

1. Clone the repository:

   git clone https://github.com/MohsenKashefi/artificial-intelligence-project.git

Navigate to the project directory: