Acknowledgments

This textbook owes its existence to the open-source data science community. Every library used in these pages — scikit-learn, pandas, numpy, XGBoost, LightGBM, SHAP, MLflow, FastAPI — represents thousands of hours of volunteer work by developers who believe that powerful tools should be freely available.

This book is written for the aspiring data scientists, the career changers, and the analysts leveling up who keep asking the same questions: "How do I actually do this at work?" "What do hiring managers really want to see?" "Why doesn't the real world look like the Kaggle dataset?"

Thank you to the open-source textbook community. The decision to release this book under CC-BY-SA-4.0 was inspired by the belief that quality education should not require a $200 textbook.

And thank you to the data scientists who have written publicly about their failures — the models that broke in production, the A/B tests that went sideways, the postmortems shared in blog posts and conference talks. That openly shared experience is what the practitioner voice in these pages tries to reflect.

This book was drafted with AI assistance and published by DataField.Dev. To the instructors and students who use it: if you find a confusing passage, a broken code example, or an exercise that is too easy or impossibly hard, corrections and suggestions are welcome (see https://datafield.dev/about.html#contact).


If you find an error, have a suggestion, or want to contribute, see CONTRIBUTING.md. This book gets better because of its community.