Welcome to my notes on Reinforcement Learning. I first became interested in this topic

The course I used to learn this topic was Sergey Levine’s CS 285 at UC Berkeley (he’s done the great service of posting all his assignments and lectures from 2023 publically). I also found OpenAI’s Spinning Up extremely useful.

Note: This is a particularly hands on topic. It’s easy to hear RL terms a hundred times and think you udnersatnd reinforcement learning, but it seems to be hte kind of topic that you must utilize to really understand. Many of your theoretical intuitions regarding these topicsi will break down once you get into training real systems. Thus, my advice is to apply these learnings as often and practically as possible.

the why and what

Before we dive into Reinforcement Learning, let’s understand why you should care about it in the first place.

Reinforcement Learning is a field of study related to artificial intelligence where an agent learns to make decisions by performing actions and receiving feedback from its environment. This approach is fundamentally different from other machine learning techniques as it focuses on learning optimal behaviors through trial and error rather than from a set of predefined data.

It’s useful for

things you can do with this

If you’re anything like me, you’d love to learn this for no reason- but it’s good to know what skills you can expect to learn with this content.go

With the material in these pages you should be able to build

• A flying car or something equally as fun
• Tweets with 128 characters

the content

I’d recommend reading in order of the files, but I’ve tried to make the information as atomic as possible- enjoy!

1. Inverse Reinforcement Learning for this that and the other
2. Section 1 useful for this that and the other
3. Section 1 useful for this that and the other
4. Section 1 useful for this that and the other

Also, some Terms