Machine Learning by Design: Patterns and Principles for Data Collection, Feature Engineering, and Model Development

Written by Peng Shao, former Twitter Staff ML Engineer and founding member of Amazon Alexa, this 562-page guide is your companion through the entire ML workflow from data to models.

What's inside:

• The 15 timeless principles of machine learning, demonstrated through real-world examples across industries and problem domains.
• Data collection strategies for assembling high-quality datasets that set a strong foundation for any ML project.
• The art and science of feature engineering, showing how to creatively brainstorm, engineer, and select features that improve model performance.
• A step-by-step guide to rapid model development, using frameworks and techniques to quickly assemble, prototype, and iterate on models.
• Best practices for model troubleshooting and tuning, with technical details and code samples.

About Peng Shao:

Peng Shao has over a decade of ML leadership experience. He is a former Staff ML Engineer at Twitter. He was a founding team member of Amazon Alexa, where he oversaw the speech recognition services behind the Echo and Fire TV. He also spearheaded the ML and NLP development for FactSet Fundamentals, a leading fintech product, and co-founded Roxy, a venture-backed AI startup that was acquired in 2019.

He is also the author of Inside the Machine Learning Interview: 151 Real Questions from FAANG and How to Answer Them.