Machine Learning Essentials
Predictive Modeling in R
Alboukadel Kassambara
A book-length lesson series
Machine Learning Essentials
Predictive Modeling in R
Classify, predict, validate, explain — the supervised models the work actually needs, with tidymodels.
26 chaptersexecutable, reproducible R
The runtime is the judge
A real figure from the book. Reproduce it yourself.
This is a genuine, reproducible figure from a chapter — not a mockup. Open the lesson, copy the code, and reproduce it in your own R session.
Where to start
Jump straight into a few key chapters, or read the whole book below.
The book is the free curriculum
There's no paywall on understanding. Every chapter is a free, runnable lesson — read the whole book online. The Pro PDF is the same content, typeset to keep.
- 01Build Models (tidymodels)
- 02Feature Engineering (recipes)
- 03Best Subsets Regression
- 04Stepwise Regression
- 05Penalized Regression
- 06PCR & PLS Regression
- 07Nonlinear Regression
- 08Decision Trees (CART)
- 09k-Nearest Neighbors
- 10Random Forest
- 11Gradient Boosting
- 12Stepwise Logistic Regression
- 13Penalized Logistic Regression
- 14Discriminant Analysis
- 15Multinomial Logistic Regression
- 16Support Vector Machine
- 17Naive Bayes
- 18Model Evaluation (ROC, Confusion Matrix)
- 19Thresholds & Calibration
- 20Accuracy Metrics (R², RMSE, AIC)
- 21Cross-Validation
- 22Bootstrap Resampling
- 23Variable Importance
- 24Partial Dependence & ICE
- 25SHAP Values
- 26Prediction Uncertainty & Trust
Every chapter is free to read; copy any block to reproduce the results.
Free online, or keep it forever
These are different goods — not a paywall on the same thing.
Free online
Every example is executable, reproducible R — copy any block and run it in your own session. The full book, online, forever free.
- The complete book, chapter by chapter
- Every example is copy-and-run reproducible R
- Always the latest edition
- No sign-up, no paywall
With Pro
- A typeset offline PDF and ePub
- Prova on your own code and data
- Verifiable certificates
- Saved work and higher compute
- All seven books, living and updated
Learn it free online. Keep it forever as a book.
You already run his code
Alboukadel Kassambara wrote the R packages a generation of data scientists reach for — ggpubr, factoextra, survminer, rstatix, ggcorrplot. More than 50 million installs. This book is the source.
50,000,000
combined package installs
From readers of the published editions.
Why Pro is worth it
The €199/year isn't for a PDF of free content. It's for Prova running on your own data, the entire continuously-updated curriculum, verifiable certificates, saved work, higher compute — and every book, living.
Less than the big subscription platforms charge for a fixed course library — and Pro runs on your own data.
Frequently asked questions
- Is the book really free to read?
- Yes. Every chapter is a free, runnable lesson online — the complete book, no paywall on understanding. Pro adds the offline typeset PDF/ePub, Prova on your own data, certificates, and more.
- Can I buy just this one book?
- The offline PDF is a Pro perk, not sold on its own. If you want a single-book, one-time purchase, the ebook is available on Kindle. Everything else lives in Pro.
- What formats does the Pro book come in?
- A typeset PDF at launch, with ePub as a fast-follow — both watermarked to your account and auto-updated from the same source as the free lessons.
- Do the books keep getting updated?
- Yes. Each book is generated from the living lessons, so your offline copy tracks the latest edition — no frozen, out-of-date PDF.
- I bought this book before — what happens to my purchase?
- Your past purchase and Kindle copy stay yours. Pro is a new, recurring option that adds the living online book plus the whole platform.
Start reading — free.
Open the book as living lessons and run every example. Keep it offline, on your data, with Pro.