I have seen this book referenced a couple of times as "the" book about how data science/ML can be applied in the security field. However, after reading it, I have to say I don't understand who is this book for. Because if you are a data scientist, this book will not be enough to get you going in the security field and vice versa, if you are a security person, you will not learn how to do data science from this book.
I appreciate the introduction; it is very well written, and succinctly states the arguments for why data-driven decision making is important for the security field. It even provides a nice quick historical context for the ML & stats fields for the layman. However, starting chapter number two, it is only worth skimming through, there is just not enough information to make it worthwhile. The most valuable are perhaps the materials referenced at the end of every chapter.