ML Graph Learning Handbook

A static handbook for Machine Learning and Graph Machine Learning notes.

Overview

ml-graph-learning-handbook is a structured collection of Markdown notes for learning classical Machine Learning, deep learning foundations, and Graph Machine Learning.

The repository includes a browser-based reader in index.html. It renders the notes as a searchable book with collapsible sections, base and expanded views, equations, Mermaid diagrams, and light/dark mode.

Table of Contents

• Features
• Quick Start
• Repository Structure
• Topics Covered
• Adding Notes
• Contributing
• License

Features

• Book-style Markdown reader
• Collapsible Overview, ML, and Graph ML sections
• Expanded notes selected by default
• Base, expanded, and all-note views
• Searchable chapter sidebar
• Previous and next chapter navigation
• Table of contents generated from headings
• MathJax rendering for equations
• Mermaid rendering for diagrams
• Light and dark mode
• Plain HTML, CSS, and JavaScript
• No install step or build step required

Quick Start

From the repository root, start a local static server:

python3 -m http.server 8000

Open the reader:

http://localhost:8000/index.html

Opening index.html directly from the filesystem may block Markdown loading in some browsers, so a static server is recommended.

Repository Structure

.
|-- index.html
|-- README.md
|-- LICENSE
|-- ML Notes.md
|-- ml/
|   |-- 001-feature-scaling-normalization-standardization/
|   |-- 002-data-preprocessing-tools/
|   `-- ...
|-- graph-ml/
|   |-- 001-graph-neural-networks/
|   |-- 002-node-edge-graph-level-features-predictions/
|   `-- ...
`-- scripts/

Each topic folder follows this pattern:

<topic>.md
<topic>-expanded.md

The base note is concise. The expanded note is more detailed and may include equations, diagrams, examples, and implementation notes.

Topics Covered

Machine Learning

• Data preprocessing
• Feature scaling
• Regression
• Classification
• Decision trees
• Random forests
• Support vector machines
• Naive Bayes
• Clustering
• Dimensionality reduction
• Model evaluation
• Ensemble learning
• Reinforcement learning
• Natural language processing
• Neural networks
• Convolutional neural networks
• Transformers

Graph Machine Learning

• Graph neural networks
• Node, edge, and graph-level prediction
• Node embeddings
• PageRank and random walks
• Message passing
• GNN expressiveness
• Heterogeneous graphs
• Knowledge graph reasoning
• Neural subgraph matching
• Community detection
• Graph generative models
• GNN limitations
• Scaling GNNs
• Graph pretraining

Adding Notes

Use numbered, lowercase, hyphen-separated folders.

For ML notes:

ml/038-new-topic/038-new-topic.md
ml/038-new-topic/038-new-topic-expanded.md

For Graph ML notes:

graph-ml/019-new-graph-topic/019-new-graph-topic.md
graph-ml/019-new-graph-topic/019-new-graph-topic-expanded.md

Recommended note structure:

# Topic Name

Short explanation.

## Intuition

## Mathematical View

## Example

## Common Pitfalls

## Summary

Use fenced Mermaid blocks for diagrams:

```mermaid
flowchart LR
    A[Data] --> B[Model]
    B --> C[Prediction]
```

Use display math for important equations:

$$
\hat{R}_{\mathrm{test}}(\hat{f}) =
\frac{1}{m}\sum_{i=1}^{m}\ell(y_i^{\mathrm{test}}, \hat{f}(x_i^{\mathrm{test}}))
$$

Contributing

Contributions are welcome. Useful contributions include:

• Fixing typos
• Improving explanations
• Adding examples
• Adding diagrams
• Adding equations
• Improving topic coverage
• Improving the static reader

Suggested workflow:

1. Fork the repository.
2. Create a branch.
3. Make focused changes.
4. Preview the reader locally.
5. Open a pull request with a clear description.

Please follow the existing note naming style.

License

This project is licensed under the MIT License.