Reinforcement Learning to Get Device Screen

This is a Python codebase that includes several classes and functions for a reinforcement learning agent that learns to navigate an environment. The agent learns to navigate from a starting position to a goal position in the grid by updating its Q-values based on the rewards obtained through interactions with the environment. Here’s a breakdown of the main components and functionalities: …

[IARTES] Reinforcement Learning to Get Device Screen!

Reinforcement learning has been making waves in artificial intelligence, enabling systems to learn optimal behaviors through interaction with their environment. The project [IARTES] Reinforcement Learning to Get Device Screen leverages Q-learning, a powerful reinforcement learning algorithm, to navigate a simulated grid environment. Here’s an in-depth look into the project, its implementation, and how you can get started.

Overview

The project demonstrates how an AI agent learns to navigate a grid environment using Q-learning. The agent starts at a defined position and learns to reach a goal by updating its Q-values through iterative training. Here’s a quick breakdown of its key components:

• Q-learning Algorithm: The core learning mechanism that enables the agent to optimize its actions based on cumulative rewards.
• Environment Simulation: A grid world environment where the agent interacts and learns.
• Visualization: Tools to track the agent’s progress, including animations and plots.

Key Features

QAgent Class

The QAgent class orchestrates the learning process:

• Training: Implements Q-learning, allowing the agent to learn optimal navigation strategies.
• Testing: Evaluates the trained model by navigating the environment using learned policies.
• Random Agent: Baseline behavior for comparison, using random actions instead of learning.

Environment

The environment simulates a grid world, offering:

• Customizable size and goal positions.
• Real-time feedback through rewards and state updates.

Visualizations

The project integrates plotting and animation utilities to:

• Visualize the agent’s learning progress.
• Create dynamic GIFs showcasing the agent’s path.

Installation

1. Clone the Repository

git clone git@github.com:raulbatalha/reinforcement_learning_to_get_device_screen.git
cd reinforcement_learning_to_get_device_screen

2. Install Dependencies

pip install -r requirements.txt

Usage

Run the main script to start training the agent:

python3 src/main.py

This will:

1. Train the agent in the grid world environment.
2. Display the agent’s path using Matplotlib.
3. Save the trained model for future use.

Project Structure

reinforcement_learning_to_get_device_screen/
├── src/
│   ├── main.py
│   ├── environment/
│   │   └── env_screen.py
│   ├── agent/
│   │   ├── qlearning.py
│   │   └── qagent.py
│   ├── utils/
│   │   └── util.py
├── data/
│   └── models/
├── media/
├── tests/
│   ├── test_environment.py
│   └── test_agent.py
├── requirements.txt
└── README.md

• env_screen.py: Implements the environment class.
• qagent.py: Handles training, testing, and visualization.
• qlearning.py: Core Q-learning algorithm implementation.

UML Diagrams

To help you navigate the code structure, here’s a UML class diagram:

Contribution Guidelines

We welcome contributions to improve this project! Here’s how you can help:

1. Fork the repository.
2. Create a branch for your feature or bugfix.
3. Commit your changes with a descriptive message.
4. Push to your branch and submit a Pull Request.

Author

Raul Batalha

• Computer Engineer
• Email
• LinkedIn

License

This project is licensed under the MIT License. MIT © Raul Batalha.

Feel free to fork, adapt, and contribute!