Machine Learning Refined

2nd Edition

Foundations, Algorithms, and Applications

Learn the foundations of machine learning, from scratch.

Explore chapter resources Open Python notebooks

Used as a reference text at 100+ universities and colleges.

Machine Learning Refined second edition book cover

The pedagogy

Visualize. Derive. Implement.

Every topic moves from geometric intuition to precise math to executable Python.

01IntuitionSee the geometry
02DerivationEarn the notation
03CodeBuild it in Python

Visual learning system

Intuitive visuals that pull you into each chapter

Source-backed figures and animations reveal the geometry, then take readers straight to the relevant chapter materials.

LookStart with a real book figure.
LearnOpen the chapter package behind it.
ExploreContinue into chapters or notebooks.

Regression as Optimization figure from Machine Learning Refined

Classification as Optimization figure from Machine Learning Refined

Exact vs Backtracking figure from Machine Learning Refined

Least Squares figure from Machine Learning Refined

PCA Geometry figure from Machine Learning Refined

Feature Transformations figure from Machine Learning Refined

Feature Learning Complexity figure from Machine Learning Refined

Regularized Feature Fits figure from Machine Learning Refined

Model Families Compared figure from Machine Learning Refined

Boundary Families figure from Machine Learning Refined

Multiclass Histogram figure from Machine Learning Refined

Unfolded Network figure from Machine Learning Refined

Ch. 2 / Optimization

Regression as Optimization

Regression fitting introduced as cost minimization.

Chapter resources Inspect image

Reviews & adoption

Trusted where machine learning is taught.

Used at 100+ universities and colleges, with professor endorsements and reader praise for its visual, first-principles approach.

For instructors See adoption proof

Professor endorsements

John G. ProakisProfessor Emeritus, Northeastern University

Highlights the book's path from basic principles to practical implementation.

Osvaldo SimeoneProfessor, King's College London

Points to the first-principles presentation, geometric intuition, and Python exercises.

David DuvenaudAssociate Professor, University of Toronto

Praises the steady build-up of tools, examples, runnable code, and detail.

Kimiaki ShirahamaProfessor, Doshisha University

Emphasizes the unified optimization viewpoint and visual explanations.

Reader praise

One of the best books in Machine Learning“What is an absolute gem are the chapters on Feature Learning, Selections and Engineering.”Reader on Amazon Gold“teaching machine learning rigorously but from first principles”Estefano Palacios on Amazon Love this book!“The content, the painstakingly created figures, and the beautiful hardback edition are all excellent.”Rama Ramakrishnan on Amazon One of the best books on the topic“Would definitely recommend to anyone starting with ML.”Julio Perez Olvera on Amazon