> For the complete documentation index, see [llms.txt](https://www.mlcompendium.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://www.mlcompendium.com/decision-intelligence/reinforcement-learning.md).

# Reinforcement Learning

A bandit decides one step at a time; reinforcement learning gives the agent a sequence of states and lets it learn by interacting for reward. The page starts with books, platforms, and lectures, then sketches Q-learning, moves to deep RL, and ends with RLHF.

The same notes are in [Contextual Bandits](/decision-intelligence/contextual-bandits.md) and [Multi Armed Bandits](/decision-intelligence/multi-armed-bandits.md).

## Introduction

The entry point is the textbook in both of its states. [Reinforcement Learning: An Introduction second edition WIP](https://web.stanford.edu/class/psych209/Readings/SuttonBartoIPRLBook2ndEd.pdf) is the second edition while it was still in progress, and the completed [book](http://incompleteideas.net/book/RLbook2020.pdf) is the finished version.

Beside the book, Vidhya has two Faizan Shaikh articles. Vidhya on [Getting ready for AI based gaming agents – Overview of Open Source Reinforcement Learning Platforms](https://www.analyticsvidhya.com/blog/2016/12/getting-ready-for-ai-based-gaming-agents-overview-of-open-source-reinforcement-learning-platforms/) explains reinforcement learning with examples and tours the major platforms such as Deepmind Lab and OpenAI Gym. Vidhya on [Simple Beginner’s guide to Reinforcement Learning & its implementation](https://www.analyticsvidhya.com/blog/2017/01/introduction-to-reinforcement-learning-implementation/) defines reinforcement learning as a type of machine learning where an agent learns to maximize reward by interacting with an environment.

The reward idea also works outside games. ZipRecruiter on [Classifying Job Titles With Noisy Labels Using REINFORCE](https://medium.com/@ziprecruiter.engineering/classifying-job-titles-with-noisy-labels-using-reinforce-ce1a4bde05e2) — Fine-grained job title classification with noisy labels using the REINFORCE algorithm and multi-task learning.

The same notes are in [Unbalanced labels](/problem-framing/label-algorithms.md#unbalanced-labels).

The reason it is here: this article has a very nice trick in adding a reward component to the loss function in order to mitigate for unbalanced class label problem, instead of the usual balancing.

For a full course, David Silver — [Home Page](https://www.davidsilver.uk/teaching/) is his Teaching page, and the Google DeepMind recordings of his RL Course are the lectures in order:

1. [1](https://www.youtube.com/watch?v=2pWv7GOvuf0): Introduction to Reinforcement Learning
2. [2](https://www.youtube.com/watch?v=lfHX2hHRMVQ): Markov Decision Process
3. [3](https://www.youtube.com/watch?v=Nd1-UUMVfz4): Planning by Dynamic Programming
4. [4](https://www.youtube.com/watch?v=PnHCvfgC_ZA): Model-Free Prediction
5. [5](https://www.youtube.com/watch?v=0g4j2k_Ggc4): Model Free Control
6. [6](https://www.youtube.com/watch?v=UoPei5o4fps): Value Function Approximation
7. [7](https://www.youtube.com/watch?v=KHZVXao4qXs): Policy Gradient Methods
8. [8](https://www.youtube.com/watch?v=ItMutbeOHtc): Integrating Learning and Planning
9. [9](https://www.youtube.com/watch?v=sGuiWX07sKw): Exploration and Exploitation
10. [10](https://www.youtube.com/watch?v=kZ_AUmFcZtk): Classic Games

The figure is from those David Silver teaching materials.

<figure><img src="https://83674056-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-Mgd48oS5_duTKOVE_Et%2Fuploads%2FLcS2KVabrePqE3KvXa0R%2Fimage.png?alt=media&amp;token=81c5d8f7-cb25-43e0-8326-270e93dbbe69" alt=""><figcaption><p>David Silver teaching materials.</p></figcaption></figure>

The last introductory source frames RL as sequential decisions in general. [Sequential Decision Analytics and Modeling](https://castle.princeton.edu/sdamodeling/) is the second edition from CASTLE, by Warren B. Powell, Professor Emeritus at Princeton University.

### Q-LEARN

With the lectures behind you, Q-learning is the first algorithm to sketch. The problem is a Markov chain problem, (state, action, new state, reward). There is lots of exploration in the beginning, then exploitation, and the method returns optimal policy. For what Q itself is, the Udacity video What is Q is on youtube [here](https://www.youtube.com/watch?v=9m_6q_KECTk).

### Deep Learning

From Q-learning, the next step is RL combined with deep learning, deep Q-learning first. [A review paper about RL in DL](https://arxiv.org/pdf/1701.07274.pdf) is Deep Reinforcement Learning: An Overview on arXiv. [deep Q-learning](https://www.analyticsvidhya.com/blog/2019/04/introduction-deep-q-learning-python/) is Ankit Choudhary's An Introduction To Deep Reinforcement Learning, which builds a deep Q-learning model in Python using keras and gym. The Pytorch tutorials then give runnable versions: [DQN](https://pytorch.org/tutorials/intermediate/reinforcement_q_learning.html) is the Reinforcement Learning (DQN) Tutorial, [PPO](https://pytorch.org/tutorials/intermediate/reinforcement_ppo.html) is Reinforcement Learning (PPO) with TorchRL, and the [Mario example](https://pytorch.org/tutorials/intermediate/mario_rl_tutorial.html) trains a Mario-playing RL agent.

## RLHF

After the RL introduction above, the same reward loop is used to tune language models from human feedback. The same notes are in [Reinforcement Learning for LLM](/generative-ai/large-language-models-llms.md#reinforcement-learning-for-llm).

[illustrated RLHF by Huggingface](https://huggingface.co/blog/rlhf) is Illustrating Reinforcement Learning from Human Feedback (RLHF), from a site that describes itself as on a journey to advance and democratize artificial intelligence through open source and open science.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://www.mlcompendium.com/decision-intelligence/reinforcement-learning.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.