Skip to main content

Introduction to MLVisual

Welcome to MLVisual, your complete platform for learning Reinforcement Learning hands-on. Here you'll find everything you need to master RL from basic concepts to advanced algorithm implementation.

🎯 What is MLVisual?​

MLVisual is an "AI Gym" that transforms Reinforcement Learning from a passive process (watching videos) to an active and practical one (training agents in interactive simulations). Our approach is based on hands-on learning through practical laboratories.

🏗️ Our Learning Approach​

Learn by Doing​

Instead of just reading theory, you'll learn by creating:

  • Simulation environments from scratch in Unity
  • Intelligent agents that learn to make decisions
  • RL algorithms implemented step by step
  • Complete projects you can showcase in your portfolio

Hub-and-Spoke Model​

Our laboratories follow a unique model:

  • Proto Labs (Hub): Learn to create simulation environments
  • RL Labs (Spoke): Train agents with different algorithms
  • Reusability: One environment serves multiple algorithms

📚 How to Use This Documentation​

For Beginners​

If you're new to Reinforcement Learning:

  1. Start with Getting Started - Set up your development environment
  2. Explore the Glossary - Get familiar with key terms
  3. Read the Theoretical Articles - Understand the fundamentals
  4. Practice with the Laboratories - Apply what you've learned

For Experienced Developers​

If you already have experience:

  1. Review the Installation Guide - Set up mlvlab
  2. Explore the Laboratories - Advanced practical projects
  3. Dive into the Articles - Advanced concepts

🎮 Practical Laboratories​

Our laboratories allow you to:

  • Create environments from scratch in Unity
  • Train agents with Q-Learning, DQN, PPO, and more
  • Experiment with different configurations
  • Visualize the learning process in real-time

Available Projects​

Ants Saga​

Our first complete project where ants learn to collect food:

  • Proto Lab: Create the environment in Unity
  • RL Lab: Train agents with Q-Learning

🧠 Key RL Concepts​

What is Reinforcement Learning?​

RL is a type of machine learning where an agent learns to make optimal decisions by interacting with its environment and receiving rewards.

Main Components​

  • Agent: The entity that learns and makes decisions
  • Environment: The world in which the agent interacts
  • Actions: The decisions the agent can make
  • Rewards: The feedback the agent receives

Why is it Important?​

RL is the foundation of many modern technologies:

  • Games: AlphaGo, OpenAI Five
  • Robotics: Autonomous navigation, manipulation
  • Finance: Algorithmic trading
  • Recommendations: Recommendation systems

🚀 Start Your Journey​

Prerequisites​

  • Basic knowledge of programming (Python/C#)
  • Motivation to learn hands-on
  • Time to experiment and practice

What You'll Need​

  • Unity 2022.3 LTS - To create environments
  • Python 3.8+ - To train agents
  • mlvlab - Our Python library
  • Patience - RL can be complex at first

💡 Tips for Success​

Learn Incrementally​

  • Start with simple concepts
  • Practice with small examples
  • Gradually advance to more complex projects

Experiment Actively​

  • Don't just follow tutorials
  • Modify parameters and observe results
  • Try to solve problems on your own

Join the Community​

  • Share your projects
  • Ask questions when you have doubts
  • Help other learners

🎯 Next Steps​

Ready to get started? We recommend following this order:

  1. Set up your environment - Install everything you need
  2. Explore the glossary - Get familiar with the terms
  3. Read the articles - Understand the fundamentals
  4. Start with Ants Saga - Your first practical project

Welcome to MLVisual! We're here to help you master Reinforcement Learning in a practical and fun way.