I was recently given a review copy of Clojure for Machine Learning. I have an academic familiarity with machine learning techniques and presented on a few at speakerconf 2012. I haven’t explored machine learning in Clojure since preparing that talk and was excited to read a book on the topic.

The book gives a shallow introduction to many different topics. It does so through a bit of mathematics and much more code. Depending on the section, the code examples implement the algorithm being discussed, show you how to use a specific library, or do both.

An aspect I particularly enjoy about the code examples is that they always start by showing what dependencies should be added to your project.clj file. This is done even if the library has been used in a previous chapter. Because of this every example can stand on its own.

Something that can almost always be improved about Clojure examples is that namespaces should be referenced using the require form with a namespace alias. Even if that require requires a namespace with a terrible alias, such as (require '[example :as e]), it makes the example code easier to understand. Being able to read e/a-func instead of a-func makes it more explicit as to where that function is located and aides understanding.

I rate all my books by the goodreads five star scale¹. This book earns three stars. Even with my limited machine learning background I didn’t learn anything new but I was introduced to some Clojure libraries and enjoyed seeing Clojure implementations of machine learning techniques.

If you enjoy Clojure and the table of contents excites you then you’ll most likely find this book interesting. If you want to maximize your learning I’d recommend taking an online course in machine learning². It will be a larger time investment but you’ll leave with a deeper understanding.