How to Solve All Machine Learning Problems

Welcome to the book Go Machine Learning Projects.

This is a rather odd book. It's not a book about how machine learning (ML) works. In fact, originally it was decided that we will assume that the readers are familiar with the machine learning (ML) algorithms I am to introduce in these chapters. Doing so would yield a rather empty book, I feared. If the reader knows the ML algorithm, what happens next is to simply apply the ML algorithm in the correct context of the problem! The ten or so chapters in this book would be completed in under 30 pages—anyone who's written a grant report for government agencies would have experience writing such things.

So what is this book going to be about?

It's going to be about applying ML algorithms within a specific, given context of the problem. These problems are concrete, and are specified by me on a whim. But in order to explore the avenues of the application of ML algorithms to problems, the reader must first be familiar with algorithms and the problems! So, this book has to strike a very delicate balance between understanding the problem, and understanding the specific algorithm used to solve the problem.

But before we go too far, what is a problem? And what do I mean when I say algorithm? And what's with this machine learning business?