Complete Guide to Unity Machine Learning Agents (Basics)

Implementation of reinforcement learning environment,
Easy and convenient with Unity!

Reinforcement learning environment ,
How can I prepare it?

Since AlphaGo made a big impact in 2016, interest in reinforcement learning , which is said to have been applied to AlphaGo, has greatly increased, and it seems that the enthusiasm is still hot. The major elements that make up this reinforcement learning are the reinforcement learning algorithm and the reinforcement learning environment, as shown below. These two exchange information such as actions, states, and rewards, and the reinforcement learning algorithm performs learning.

Since AlphaGo, reinforcement learning algorithms have made a lot of progress. Accordingly, various types of reinforcement learning environments such as OpenAI GYM, Mujoco, Atari, GTA5, Malmo, etc. have also been released. Most of these environments are game-based. Reinforcement learning is clearly a good algorithm to apply to games, but recently, attempts to apply reinforcement learning to various fields such as recommendations, robots, drones, energy, and finance are increasing.

However, reinforcement learning environments for these various fields are still lacking. In particular, it is very difficult to expect that an environment that precisely satisfies the specific specifications desired by developers will be disclosed. Even if there is a robot environment with a specific sensor configuration and joint structure that you want to apply reinforcement learning to, it may be impossible to even start research if there is no public reinforcement learning environment for that field.

But in September 2017, Unity, one of the world's largest game engine companies, released a tool called Unity Machine-Learning Agent that can solve this problem.

---

With Unity ML-Agents
Implementing reinforcement learning environment!

In this lecture, you will learn how to implement various reinforcement learning environments using this Unity machine learning agent, as well as the theory and code implementation of reinforcement learning algorithms that can be applied to the environments.

Information before taking the class!

The content of this lecture contains the same content as the book "Learning Reinforcement Learning with PyTorch and Unity ML-Agents" below! Please be aware of this before taking the course!

Complete Guide to Unity Machine Learning Agents - Basics

The entire content of the Unity Machine Learning Agent Complete Mastery lecture will be divided into the basics and application parts, and this lecture will cover the basics part. The specific content to be covered in the basics part is as follows.

• Reinforcement Learning Basics Terms and Theory
• Unity Installation and Basic Usage
• Unity Machine Learning Agents Installation, Components Description, Usage (mlagents-learn, Python API)
• Environment Creation
  • GridWorld, Drone, KartRacing
• Learning reinforcement learning algorithm theory and implementing code
  • DQN, A2C, DDPG, Behavioral Cloning

The code for the environment we will create and the algorithms we will learn in this lecture are all included on GitHub .
The images below are the reinforcement learning environments you will implement in this lecture and the results of learning using the reinforcement learning algorithm you will implement.

_{Creating a Gridworld Environment}

_{Creating a drone environment}

_{Creating a kart racing environment}

---

Frequently Asked Questions
Check it out.