RL Starts Here: Setting the Scene
Markov Decision Processes (MDP)
Loading notes…
One Oz of Humanity.
Here, you will find my notes on things I read or learnt.
What is this website?Entries
Reinforcement Learning is a fascinating topic. Nowadays with the help of AI, it takes little effort to set up an RL experiment for most general applications. But the results always hit a bottleneck: either when the environment gets more complex, or the requirements change. To understand it properly, I started "relearning" RL from the ground up, interpreting it in my own words, one concept at a time.
About
This is a personal learning project.
I have been searching for the best way to learn, especially the obscure, theoretical stuff. I could go get a master's degree, or, I think, a better alternative: taking advantage of AI, while at the same time ensuring the knowledge is going into my brain, and preventing me from devolving into a meat proxy.
I started writing these notes because I miss the handwritten-note-feeling. I also know that the best way to learn is to teach. The increased friction from doing it by hand helps me comprehend and retain the knowledge. And knowing that someone might be reading it forces me to be more concise and mindful of the things I write down.
Hence the title of this website, "One Oz of Humanity." I consider this my "one oz effort" to offer back a piece of humanity to the AI-infiltrated internet.
Human reader, a couple of heads-up:
- All the content here is completely handwritten/typed.
- I even left my typos in there to keep you on your toes.
- For things I copied, referenced, quoted, or AI generated, I will highlight them like this.
- This webpage was thrown together using an agent. But I come up with the design, the picture here is proof.
- This is a work in progress.
