The book of Goodfellow et al. is about neural networks. It's not more complete than a book on machine learning.
Maybe it’s because of the 100 page limit, but it’s pretty unusual to put supervised linear methods with unsupervised clustering. Usually, those are two different chapters.