A few weeks ago, I posted a short message on Facebook saying that I was going to write a new Machine Learning ebook. I have told my idea and chatted to a few people since (engineers, hackers and bootstrappers), and I am glad to have confirmed that there is indeed a need to be addressed with this ebook.
The most prominent question I have had during these chats simply was:
What is Machine Learning (ML)?
I could give a mathematical answer, but that would probably have zero impact on you… More interesting are the questions that would follow: “What can you do with Machine Learning? What are the limits? How does it relate to other Artificial Intelligence techniques? Can I use Machine Learning in what I do, for my particular problem? How?”
Actually, the 2010s have seen ML advance to a point where anyone (that means you!) can use it to produce meaningful results, without having to understand the maths and theory on which ML is based. ML algorithms are now exposed through extremely simple interfaces, so you can effectively do predictions thanks to Machine Learning “black boxes”, without knowing the workings in details. Sure, you will need to understand a few basic concepts and what the algorithms are supposed to do.
The focus of my ebook is going to be on teaching you just enough to bootstrap Machine Learning in your projects/hacks/business. Obviously, on a technical level you won’t get as much from ML as if you’d hire an expert. But you don’t have the resources anyway, and you don’t even know yet what use ML is going to be of. The way to go in that case is to bootstrap! Experiment with your new ML superpowers, validate ideas, and create prototypes. You may decide in the future to optimise things by hiring or becoming an expert, and you will then be in a much better position to do so. Or you may stay happy with your ML setup and choose not to change anything — basic algorithms can work wonders.
Why the ebook form? Because it’s intended to be a much quicker read than other ML resources for newbies. I am going to teach you how to use ML APIs instead of how to program ML algorithms. This is going to abstract the scientific and algorithmic complexity, so you can read the ebook and start putting what you’ve learnt to practise in just a day. By the way, I intend to write a hands-on chapter with a concrete use case and code to help you get real!
If there’s anything that you would like to see in this ebook, now’s your chance to speak your mind. You can let me know in the comments what you would like to learn about, or you can just ask me anything. Also, if you like/share this post and give me shout, you’ll get a chance to receive a draft version of the ebook for free (pricing is going to be somewhere around $50).