Book review: Effective Amazon Machine Learning

All major cloud providers provide some support for Machine Learning algorithms. They also evolve all the time. There are not many books ont he subject, due to the evolution of these services, so let’s have a look at this one.


I actually started a long, traditional review, but couldn’t really go on. The main reason is that the book targets a very small part of the current AWS offering on Machine Learning. Then, what is odd is that the author makes a difference between machine learning and statistical methods. Actually, the latter are part of the former. No, machine learning is not about black box. That’s almost only neural networks, and even there, people are trying to figure out explanations. Linear regression is not a black box, nor are decision trees.

Decision makers need explanations, so this is just inconsistent. The other issue is the interpretation of the Gaussian distribution for linear regression (the only algorithm available in AML). Despite not introducing statistics, the author is still using a statistical interpretation, after rejecting statistical methods as machine learning!!, for the least square errors.

Also, the author keeps on confusing SGD (Stochastic Gradient Descent), the optimization algorithm, with linear regression, the model.

The good parts are the preprocessing steps, but then the AWS training sessions can cover them.


The book has some very bad bias, I’m not 100% sure the author understands the broad picture.

The main issue with the book is that despite it presents preprocessing steps that are still in use today, the main content of the book should have been Amazon Sagemaker, which does more or less anything in Amazon Machine Learning.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.