Runnable from page one
Every example lists its dependencies in a single pip line and is heavily commented, so you can open a chapter and execute it immediately.
Code-first · Math-light
Learn machine learning by building it.
A practical, self-contained guide that teaches ML through runnable Python, real datasets, and end-to-end case studies, not proofs and derivations.
930+ pages
17 chapters
90+ runnable examples
scikit-learn · PyTorch · TensorFlow
5
parts, foundations to capstone
7+
applied case studies
100%
self-contained, run as you read
CSC 373
university course adoption
Why this book
Built for people who learn by doing.
Most introductions start with the math and hope the intuition follows. This one starts with intuition and working code, then explains every step so nothing feels like a black box.
Real, openly licensed data
Work with genuine datasets, fraud, spam, sensor signals, traffic signs, financial time series, not toy numbers, all properly cited.
Ethics woven in
Fairness, bias, and responsible-use notes appear throughout, so you learn to weigh accuracy against accountability from the start.
What's inside
A complete path across five parts.
I
Foundations
What machine learning is, the Python and workflow groundwork, and how to think about a problem end to end.
II
Supervised Learning
Linear and tree-based models, then rigorous evaluation, calibration, and model selection without data leakage.
III
Unsupervised Learning
Clustering, dimensionality reduction, and visualization for finding structure in unlabeled data.
IV
Advanced Methods & Modern Applications
Neural networks in PyTorch, convolutional networks for images, interpretability, and model monitoring.
V
Capstone & The Road Ahead
An integrated capstone framework plus a forward look at transformers and foundation models.
Case studies throughout
Concepts land through applied projects you can adapt to your own work.
Applied case studies
+House price prediction
+Credit-card fraud detection
+Spam classification
+Human-activity recognition (PyTorch)
+Traffic-sign recognition (CNNs)
+Algorithmic fairness in recidivism
+End-to-end NVIDIA stock forecasting
Classroom-proven
Adopted as required reading for CSC 373, Machine Learning.
Selected for undergraduate instruction at the University of Advancing Technology for its practical, student-centered approach, and recommended for other universities and instructors seeking a hands-on machine learning text.
Who it's for
If you can write basic Python, you're ready.
Students
Undergraduates who want a practical path into ML without a heavy math prerequisite.
Career-switchers
Software engineers and analysts moving into applied machine learning work.
Instructors
Educators who need a reproducible, lab-ready text for a hands-on course.
Start today
Open Chapter 1 and run your first model.
Everything you need is in the book, code, datasets, and explanations included. No scattered repos, no missing pieces.