The Hundred-Page Machine Learning Book

Not so much a summary, but some thoughts:

1. Understanding the theory of how machine learning algorithms work is essential for data scientists.
2. Understanding the math enables a data scientist to go further - to design custom algorithms, which is increasingly important for solving complex problems.
3. Machine learning isn’t all about algorithms. Feature engineering is arguably more important.
4. Don’t stop at model evaluation and recommendation. Operationalise your models by learning how to deploy and scale them for effective and pervasive use.