1st Edition

Model-Based Machine Learning

By John Winn Copyright 2024
468 Pages 179 Color & 77 B/W Illustrations
by Chapman & Hall

468 Pages 179 Color & 77 B/W Illustrations
by Chapman & Hall

Today, machine learning is being applied to a growing variety of problems in a bewildering variety of domains. A fundamental challenge when using machine learning is connecting the abstract mathematics of a machine learning technique to a concrete, real world problem. This book tackles this challenge through model-based machine learning which focuses on understanding the assumptions encoded in... Read more

Introduction. How Can Machine Learning Solve my Problem? 

1. A Murder Mystery 

2. Assessing People’s Skills 

Interlude. The Machine Learning Life Cycle 

3. Meeting Your Match 

4. Uncluttering Your Inbox 

5. Making Recommendations 

6. Understanding Asthma 

7. Harnessing the Crowd 

8. How to Read a Model  Afterword

Biography

John Winn is a Principal Researcher at Microsoft Research, UK.

"I read Model-based Machine Learning with enthusiasm and curiosity, finding it well writtenand captivating. This book helps understand some of the most common models and it can beused by both beginners and more experienced individuals. I particularly appreciated that itpushes readers to think critically about the assumptions, sometimes hidden, their validity, andtheir consequences on the results obtained. As people’s lives are influenced by machine-madedecisions, understanding the assumptions that affect the behavior of a a machine learningalgorithm and making sure that it is transparent, interpretable and fair is an essential skill tohave for anyone who is responsible for any stage of a model’s development."
-Emanuela Furfaro, in Journal of the American Statistical Association, October 2024