Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Mastering machine learning with Python in six steps : a practical implementation guide to predictive data analytics using Python
Swamynathan M., Apress, New York, NY, 2017. 358 pp. Type: Book (978-1-484228-65-4)
Date Reviewed: Oct 18 2017

In 1969, there was a humorous film in theaters called If It’s Tuesday, This Must Be Belgium. It was about a group of American tourists taking an 18-day guided tour of nine European countries. The premise was so compelling that the movie was remade for TV in 1987, and the title alone brings a smile to one’s face as it seems to be an archetype of human experience. We often try to cram so much into so little time that we end up missing the point of everything. And this archetype is the only way I could think of to describe Mastering machine learning with Python in six steps. This book is densely packed with information about Python programming features and approaches to machine learning, much of which is recognizable but very little of which is comprehensible if one does not already know it.

The book begins in chapter 1 with a very compressed review of the Python programming language. One cannot possibly learn Python from this chapter, but the author does not make any claim to that effect. According to the author, the book is for “Python programmers or data engineers looking to expand their knowledge.” Fair enough! But the “or” should be an “and” as any shortcomings in either of these areas will create a substantial uphill climb for the reader. Personally I like these dense summaries of the language, as I nearly always come across something useful that I didn’t already know. And this chapter was not a disappointment in that regard.

Chapter 2 turns to machine learning and provides a series of paragraphs with headings such as “Statistics,” “Data Mining,” “Data Analytics,” “Descriptive Analytics,” “Diagnostic Analytics,” “Predictive Analytics,” “Prescriptive Analytics,” and “Data Science,” each with a half to a third of a page description. The topics just mentioned are covered in six pages before moving on to more headings and paragraphs. Along the way is an unreadable graphic of machine learning categories followed by more headings and paragraphs. The chapter wraps up with code examples from Numpy and Pandas, which you must already know in order to understand the examples. Whew, nine countries in 18 days! The tour described above will work if your goal is to have a picture of yourself at the Parthenon without actually knowing what the Parthenon is. And the topics in this chapter will work if your goal is to say you read something somewhere about a topic, but really don’t know anything about it.

As we proceed with the following chapters, the material gets deeper with the explanations and the code gets slightly longer, but never enough to really explain what is going on. In short, the book is way too ambitious on the range of topics and way too inadequate in explaining them.

Why six steps? Well, there are six chapters although it is hard to see them as a step progression. The author offers the following explanation: “The six-step path has been designed based on the ‘six degrees of separation’ theory that states that everyone and everything is a maximum of six steps away.” Most people would recognize this as the popular game Six Degrees of Kevin Bacon, which asserts that there are at most six degrees of separation between any one person on the planet and any other. How this idea relates to the structure of the book and its six chapters is a mystery unless the goal of the book is to keep the reader separated from the material.

The sad thing about this largely unreadable book is that the author appears to be quite knowledgeable even if that knowledge eludes the reader. It is not easy to reduce these topics down to a page or less, and although I did not read every single one, I did not find any mistakes in the ones I did read. I tried to figure out what would motivate someone to write something this ambitious that clearly did not have the reader in mind. The only thing I could think of was that it may have been written originally as a comprehensive exam for a doctoral program where the author is attempting to demonstrate proficiency before moving on to doctoral research. I could be wrong about that, but it is the only explanation that makes sense to me. And if I am wrong, the book itself is as much of a mystery as the six degrees of separation.

Having said all that, in fairness, I should also add that if you are the kind of person who refers to documentation or written explanations only as a last resort, and prefers to work through code samples in order to understand something, then this is the perfect book for you.

More reviews about this item: Amazon, Goodreads

Reviewer:  J. M. Artz Review #: CR145597 (1712-0786)
Bookmark and Share
  Editor Recommended
Featured Reviewer
 
 
Learning (I.2.6 )
 
 
Python (D.3.2 ... )
 
Would you recommend this review?
yes
no
Other reviews under "Learning": Date
Learning in parallel networks: simulating learning in a probabilistic system
Hinton G. (ed) BYTE 10(4): 265-273, 1985. Type: Article
Nov 1 1985
Macro-operators: a weak method for learning
Korf R. Artificial Intelligence 26(1): 35-77, 1985. Type: Article
Feb 1 1986
Inferring (mal) rules from pupils’ protocols
Sleeman D.  Progress in artificial intelligence (, Orsay, France,391985. Type: Proceedings
Dec 1 1985
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy