Skip to content

Data Science Field Guide

Welcome scientists, engineers, mathematicians, and curious learners.

This field guide is a compact collection of practical formulas, intuitive notes, and visual insights for the core concepts of data science and machine learning. Use it as a quick reference to refresh your memory and support clear thinking whenever these ideas appear in your work.

A Markdown-first field guide for data science and machine learning concepts, organized as flashcards.

Live site: https://rollingstorms.github.io/data-science-field-guide/

What is this?

A compact reference of concept cards covering the math, statistics, and ML topics that come up in real data science work.

Each card is a short, structured page with definitions, formulas, intuition, examples, and links to related concepts.

Who is it for?

  • Data scientists who want a fast refresher while working
  • ML engineers who need quick concept lookups
  • Students moving from theory to practical ML
  • Anyone building a durable personal/reference knowledge base for data science

Why does it exist?

To make review and recall fast.

Instead of searching scattered notes, docs, and textbooks, this project keeps commonly used concepts in one consistent format that is easy to browse, extend, and revisit.

How is it organized?

  • cards/: topic cards grouped by category (for example probability/stats, machine learning, metrics, linear algebra, optimization, graphs, NLP, deep learning)
  • index/: generated index, tags, and glossary pages
  • assets/: generated metadata plus UI helpers/styles for the site
  • scripts/build_index.py: generates indexes, tags, metadata, and mkdocs.yml from card frontmatter
  • site/: built static site output

Card categories

Current card categories under cards/:

  • activations/
  • calculus/
  • deep-learning/
  • graphs/
  • info-theory/
  • linear-algebra/
  • machine-learning/
  • ml-metrics/
  • nlp/
  • optimization/
  • probability-stats/
  • signal-processing/

How do I use it?

Browse (fastest)

  • Read cards directly in cards/
  • Or open the built site in site/ (after a build)

Build / serve locally

Generate indexes/tags and run MkDocs locally:

./run.sh

Or run specific MkDocs commands:

./run.sh build
./run.sh gh-deploy

Dynamic index/tags + MkDocs

Everything is generated from card frontmatter:

./run.sh

This runs: - python scripts/build_index.py (regenerates index/INDEX.md, index/TAGS.md, and mkdocs.yml) - mkdocs serve (or any mkdocs subcommand you pass)

Examples:

./run.sh build
./run.sh gh-deploy

Requirements

pip install -r requirements.txt

Style + spec

  • Canonical card spec: SPEC.md
  • Authoring template: _template/card.md
  • Visual style rules: _template/style.md