Synopsis: Covers the theory and application of analytics specifically in business applications.
I’ll just jump straight to the point: Provost and Fawcett have produced one of the best all-around books on analytics for business I’ve read so far.
What makes it so good? The authors start out simply enough with introductory explanations of several key ideas, but by the end of the book one has been exposed to enough theory to understand the “why” of models, not just the “how”. They don’t spare readers the math involved, but they do go the extra step to make equations more readable by eliminating luxury items like sigma and hat notations.
There is in-depth discussion of several real-life examples, including the now famous (infamous?) Target pregnancy-predicting case. I was very disappointed that Provost and Fawcett chose not to address the ethical implications of Target’s actions, choosing instead to opt out of that discussion:
Target was successful enough that this case raised ethical questions on the deployment of such techniques. Concerns of ethics and privacy are interesting and very important, but we leave their discussion for another time and place.
This is an important issue, but nearly every author in Data Science is either giving it a cursory examination or avoiding the topic altogether. Are we going to be self-regulating in this regard, or are we going to invite government intervention because of our inaction?
Other than this oversight, Data Science for Business is an excellent book. Readable, with the right balance of theory and practical application, it hits the spot without feeling bloated. And when you think about it, that is exactly what a business analytics practitioner is looking for.