Skip to main content

"Artificial Intelligence and Machine Learning for Business" by Steven Finlay

I've been promising for some time to put together some reviews of books related to programming, AI, networking, and other topics I tend to talk about on this blog. Well, I've finally been able to sit down and read a book (based in no small part to the fact that I was on a plane last week).

Yep. It seems that one of the few times I really get an opportunity to read books is when I travel.

This week I had an opportunity to read "Artificial Intelligence and Machine Learning for Business: A No-Nonsense Guide to Data Driven Technologies" by Steven Finlay.



(Full Disclosure: Links to Amazon in here use my affiliate ID. It doesn't change your price for the book, but if you do buy through one of the links I get an extremely small commission.)

It's the perfect read for a plane. In fact, the author states right up front in the forward that this edition is intended to be read in 112 minutes. He was remarkably accurate, I completed the book in approximately 120 minutes on a plane while being interrupted by my airborne neighbors requesting drinks, standing in the isle and chatting with other passengers (even though they aren't supposed to - a lapse in judgement possibly brought on by the drinks), and having to go to the lavatory (again - lots of drinks).

Mr. Finlay's entire focus in the book is not the technical details of machine learning and AI, it's the application of this technology to business. It's a recurring and refreshing theme in the book.

Every machine learning project should start with some basic questions, including "What is the business proposition being addressed?" It's an important question to ask, and is something that people from academic backgrounds or who are fresh out of school need to consistently ask themselves. While many AI projects might be cool or interesting or state-of-the-art, if you work for a company that doesn't have a lab dedicated to fundamental research, every project needs to address a very specific business case.

The book has a natural progression, from deciding whether your business is ready for prescriptive analytics, either in a guided or automated fashion, what types of questions can be reasonably answered with machine learning approaches (and, importantly, the difference between regression and classification problems), and - finally - using these approaches and integrating them into your business processes.

Those with an established technical background in machine learning will benefit from following the author's "business case first" reasoning, while the business owner/manager will benefit from the careful explanation and definition of the terms, concepts, and techniques used by data scientists, and what types of limitations the resulting models have.

I think it's a worthwhile read, whether you are enjoying a comfortable and quiet weekend afternoon, or trying to ignore a slightly over raucous group of companion travelers on a plane.

Popular posts from this blog

Neural Network Dense Layers

Neural network dense layers (or fully connected layers) are the foundation of nearly all neural networks. If you look closely at almost any topology, somewhere there is a dense layer lurking. This post will cover the history behind dense layers, what they are used for, and how to use them by walking through the "Hello, World!" of neural networks: digit classification.

Arrays of Structures or Structures of Arrays: Performance vs. Readability

It's one of those things that might have an obvious answer if you have ever written scientific software for a vector machine. For everyone else, it's something you probably never even thought about: Should I write my code with arrays of structures (or classes), or structures (or classes) of arrays. Read on to see how both approaches perform, and what kind of readability you can expect from each approach.

Genetic Algorithms: Mutation

So far in my series on Genetic Algorithms I've covered:  Basic Concepts ,  Encoding ,  Selection , and  Crossover . Those operations will get you most of the way to a functioning optimizer. However, all of the operations we've discussed exert convergence pressure  on the population. That is, selection and crossover have a tendency to draw the population together. The last operator I want to talk about -  mutation - is different. It wants to split the population apart. And as counter intuitive as it seems,  divergence  can be even more important than convergence.