MATLAB and Simulink Training

Reinforcement Learning Onramp

Reinforcement Learning

This free, two-hour tutorial provides an interactive introduction to reinforcement learning methods for control problems.

Prerequisites: MATLAB Onramp

Details and launch

 

Access to MATLAB through your web browser

 

Engaging video tutorials

 

Hands-on exercises with automated assessments and feedback

 

Lessons available in English only

Select a Lesson to Get Started

1.

Overview of Reinforcement Learning

Familiarize yourself with reinforcement learning concepts and the course.

  • What is reinforcement learning?
  • Course overview
  • Simulating with a pretrained agent

2.

Defining the Environment

Define how an agent interacts with an environment model.

  • Components of a reinforcement learning model
  • Defining an environment interface
  • Rewards and training
  • Including actions in the reward
  • Connecting a Simulink® environment to a MATLAB agent

3.

Defining Agents

Create representations of reinforcement learning agents.

  • Critics and Q values
  • Representing critics for continuous problems
  • Creating neural networks
  • Actors and critics
  • Summary of agents

4.

Training Agents

Use simulation episodes to train an agent.

  • Training
  • Improving training
Details and launch

Related Courses

Deep Learning Onramp

Get started quickly using deep learning methods to perform image recognition.

Simulink Onramp

Get started quickly with the basics of Simulink.