The book is ok to get some level of overview of some machine learning practices. It doesn't really explain much (mathematical) background of what's going on, so if you don't have that background already you will have a harder time in places. The examples feel a little random, but on the other hand they cover a fairly broad spectrum of application areas which is good. Probably the most important issue with the book nowadays is that it was written in 2017, and in the fast evolving landscape of ML, especially deep learning model architecture, it often feels woefully out of touch with topics of interest today. Also, a lot of the Jupyter notebooks don't run anymore since the software APIs of the current versions of tools used in them have made changes that are not backed compatible. It would have helped if the authors had provided a docker image (or similar) with frozen versions of all the Python libraries that the Jupyter notebooks depend on.