Reinforcement Learning Series

Bite-Sized, Self-Paced, & Practical.

Feeling guilty for not completing any online course? Each course contains 5 to 8 small tutorials in multi-media format. Each tutorial is designed to be completed in 10-15 minutes. We take a Top-Down design approach to make things intuitive. We use Agile to build fast. We use storytelling to make things fun.

Here is a sample curriculum:

1. Introduction, Conceptual Design, & Key Building Blocks

2. Build the MVP Skeleton and see how it work

3. Build the Environment

4. Build the “Brain” of the RL

5. Complete a mini-challenge and show it off in your portfolio

Build First, Theory Follows.

Overwhelmed by theories and didn’t end up building anything? There are lots of theories about RL online, but not enough re-usable code. We provide just enough clean, simple, and working code for you to get started. Never not finish a project, again. You can even re-use the code for future RL projects.

Here is an example of the code structure:

• environment.py

• RL_brain.py

• run.py

This screenshot is captured from the actual code. Each tutorial includes solution code with full documentations.

Portfolio Ready.

Not sure if your work is good enough to share? Each tutorial includes a set of mini-challenges. Each challenge aims to provide guidance for you to build out an impressive portfolio. You can show it off to your friends, colleagues, and potential employers.

Here are some sample challenges:

• modify the code to allow your RL agent to learn 2x faster

• plot the changes of reward as the RL agent learns

• develop an API to serve your solution to other users

As a data science interviewer, I have to say the best way to impress is to show off your projects. An action is louder than a thousand words, a portfolio is better than a resume, especially in the crowded ML space.

Clean, Multi-Media, and Intuitive Content

Available Tutorials