Mid-Summer Break, Open AI Gym Series!

We're taking a little bit of a mid-summer break from new content here at MMH. But we have done some extra work in organizing the site! Last week we wrapped up our series on Haskell and the Open AI Gym. We've now added that series as a permanent fixture on the advanced section of the page!

Here's a quick summary of the series:

Part 1: Frozen Lake Primer

The first part introduces the Open AI framework and goes through the Frozen lake example. It presents the core concept of an environment.

Part 2: Frozen Lake in Haskell

In the second part, we write a basic version of Frozen Lake in Haskell.

Part 3: Blackjack

Next, we expand on our knowledge of games and environments to write a second game. This one based on casino Blackjack, and it will start to show us common elements in games.

Part 4: Q-Learning

Now we start getting into the ideas of reinforcement learning. We'll explore Q-Learning, one of the simplest techniques in this field. We'll apply this approach to both of our games.

Part 5: Generalized Environments

Now that we've seen the learning process in action, we can start generalizing our games. We'll create an abstract notion of what an Environment is. Just as Python has a specific API for their games, so will we! In true Haskell fashion, we'll represent this API with a type family!

Part 6: Q-Learning with Tensors in Python

In part 6, we'll take our Q-learning process a step further by using TensorFlow. We'll see how we can learn a more general function than we had before. We'll start this process in Python, where the mathematical operations are more clear.

Part 7: Q-Learning with Tensors in Haskell

Once we know how Q-Learning works with Python, we'll apply these techniques in Haskell as well! Once you get here, you'd better be ready to use your Haskell TensorFlow skills!

Part 8: Rendering with Gloss

In the final part of the series, we'll see how we can use the Gloss library to render our Haskell games!

You can take a look at the series summary page for more details!

In a couple weeks, we'll be back, this time with some fresh Rust content! Take a look at our Rust Video Tutorial to get a headstart on that!