Course Information
Registration period
Year-round Recruitment
Price
-
Course Level
Study Mode
Duration
21 Hour(s)
Language
Cantonese, English
Location
-
8 views
Course Overview
Introduction
This course gives you an interesting and fun way to get started with reinforcement learning (RL) on DeepRacer. RL is an advanced machine learning (ML) technique that takes a very different approach to training models than other machine learning methods. Its super power is that it learns very complex behaviors without requiring any labeled training data. Our hands-on labs can help you learn , start training reinforcement learning models and test them in an exciting, autonomous car racing experience.
Time: 9:30am – 12:30pm + 1:30pm to 4:30pm HKT
Duration: 21 hours
Mode: Online + Classroom
What You’ll Learn
Course Outline:
Introducing the Game
- Why DeepRacer?
- About this Course
Module 1: How It Works
- Reinforcement Learning
- DeepRacer and AWS Services
- Action Space
- Discrete or Continuous Action Space
- Reward Function
- Reward Function Example
- HyperParameters
- Lab: Creating Your 1st DeepRacer Model
- Lab: Evaluating Models with Simulation
Module 2: Advanced Reward Function
- Cloning a Trained Model and Starting a New Training Pass
- Input Parameters of the AWS DeepRacer Reward Function
- Lab: Advancing the Model with Custom Reward Function
- Lab: Evaluating the new Model
Module 3: WayPoint Racing
- What is WayPoint?
- Lab: Preparing WayPoints for your Agent
- Lab: Creating the ML Model for WayPoints
- Lab: Racing Line with WayPoint
- Lab: Evaluating the WayPoint Model
Module 4: Advancing the Model with Log Analysis
- Training Logs of Amazon SageMake and RoboMaker
- JupyterLab and Jupyter Notebook
- DeepRacer Log Analysis Tool
- Bonus Lab: DeepRacer Training Log Analysis
