Build a solid foundation in math, CS, and AI through an intuitive open-source guide
If you want to understand the essence of concepts hidden behind complex formulas, this open-source textbook provides a path. It weaves mathematics, computer science, and artificial intelligence into a single flow, covering topics from fundamentals to practical application. It allows you to systematically build the core knowledge required to become an AI researcher or engineer.
Moving beyond the limits of existing textbooks
Most textbooks obscure good ideas with dense symbols, omit intuition, and assume the reader already knows half the material. In the rapidly changing field of AI, this approach easily becomes outdated. This book is written for curious practitioners to understand topics deeply from the ground up. It focuses on truly grasping concepts rather than just passing exams or interviews.
A knowledge map composed of 18 chapters
It starts with linear algebra basics like vectors and matrices, then covers calculus, statistics, and probability.
- Mathematical Foundations Explains vectors, matrices, calculus, statistics, and probability intuitively.
- Core AI Fields Covers machine learning, NLP, computer vision, speech processing, and multimodal learning.
- Systems and Infrastructure Includes operating systems, data structures, GPU programming, and ML system design.
How to read with an AI assistant
The repository includes an MCP server, allowing various AI assistants like Claude Code, Cursor, and VS Code to use this textbook as a knowledge base. You simply need to clone the repository locally and run the server. Educational tools and example implementations are provided so you can verify concepts directly in code.
Advice for learning
The author recommends daily cumulative self-study after classes and a 'shadow reading' technique before exams, where you explain concepts by looking only at slide titles. This method is similar to masked language modeling in machine learning, where you re-read only the missed parts and implement them in code to build muscle memory. This approach was also effective for friends who lacked confidence.
Prerequisites before reading this book
You only need elementary math knowledge and basic Python programming skills. The rest can be acquired naturally as you read. The author states that this book is particularly useful for diligent learners who are not geniuses, like 'Darwin.' It is suitable for learners who value the quality of knowledge consumption and the intensity of execution, rather than a system for those who learn quickly.
