This project aims to develop an autonomous agent capable of navigating through the Empty-Grid 6x6 environment. It encompasses recording observations and actions, building and training a Convolutional Neural Network (CNN), and evaluating the model's performance.

The project is structured into three main components:

1. Recording Observations and Actions: Utilizing the record.py script, we manually navigate the agent through the MiniGrid environment to collect data. This data consists of observations (environment states) and the corresponding actions taken by the agent.

2. Building and Training the Convolutional Neural Network: The training.py script outlines the process of constructing a CNN model. This model is trained on the recorded observations and actions, learning to predict the best action based on the current state of the environment.

3. Model Evaluation: After training, the model's performance is assessed using the test.py script. This involves running the trained model in the MiniGrid environment to evaluate its ability to navigate autonomously.

Ensure you have the following dependencies installed:

• gymnasium==0.29.1
• minigrid==2.3.1
• numpy==1.26.4
• pygame==2.5.2
• scikit_learn==1.5.0
• tensorflow==2.16.1

You can install all required dependencies by running:

pip install -r requirements.txt

To start recording observations and actions, run:

python record.py

Navigate the agent through the environment using keyboard controls. The observations and actions will be saved for training.

To train the CNN model with the recorded data, execute:

python training.py

This script will preprocess the data, build the CNN model, and train it.

To evaluate the trained model's performance, run:

python test.py

This will use the trained model to autonomously navigate the MiniGrid environment, and output the performance metrics.

Contributions to this project are welcome. Please feel free to fork the repository, make changes, and submit pull requests.

This project is licensed under the MIT License - see the LICENSE file for details.