What is Machine Learning and Why Now is the Best Time to Learn It

Introduction to Machine Learning

Machine Learning (ML) is one of the most impactful technologies that have (re)emerged in the last 10 years. Driven primarily by the exponential growth of data produced by people using many internet-related technologies and, more recently, by objects that interact with our environment (the Internet of Things) which are constantly collecting and sending data to massive storage centers, which have been a fundamental part of the expansion that ML has had in recent years. After all, today we have enough computing power at an affordable price to extract patterns from massive data sets (Big Data) and obtain valuable information from them.

Simplifying things, we can say that ML is a term that refers to a set of algorithms that learn using past data to create predictions, simulations, or analyses. These algorithms, once programmed, are represented as decision trees, neural networks, etc. Ok, that definition wasn’t very academic, so I’ll add two more accurate ones below.

Arthur Samuel (1959) - Machine Learning: Field of study that gives computers the ability to learn without being explicitly programmed.

Yes, that definition isn’t very new, but in the field of computer science, what is? We’re constantly sold ideas as novel when they were actually invented or theorized long ago. But for a somewhat more modern (and more technical) definition, here’s this one:

Tom Mitchell (1998) - Well-posed Learning Problem: A computer program is said to learn from experience E with respect to some task T and some performance measure P, if its performance on T, as measured by P, improves with experience E.

I know. It sounds like a riddle, but wait, before you leave this post and say that Machine Learning is only for geeks, nerds, or crazy people with too much free time. I can guarantee that with a little knowledge of the subject it will make sense to you, and below I’ll detail why and the use of ML in the real world.

Machine Learning Applications

ML applications already affect us all in one way or another. From the search engine we use to the facial recognition in our modern cameras, not to mention the tailored recommendations that many e-commerce services like Amazon make to us every day. ML is everywhere. And the trend is that even more applications will be found in the future than are currently known.

How and Where Do I Learn About Machine Learning?

The first thing is to really have the motivation to learn more about the subject. I want to keep this post brief, but I’m not satisfied with how little I’ve delved into the benefits of learning some ML. However, this won’t be the last post on the subject. So you can be sure that week by week you’ll find something new on the blog related to ML and, above all, practical examples, guides, and tutorials that show you the true potential of ML. But of course, I’ll leave you first with a list of what I consider the best resources to learn and delve deeper into the subject, and best of all, they are completely free resources, but of enormous quality.

Recommended Resources

1. Mathematical Foundations: You’ll surely need to review or learn some mathematics, especially linear algebra and some integrals and derivatives. For this, I recommend a series of courses from the Polytechnic University of Valencia that covers everything you need, which you can find here.

2. Andrew Ng’s Course: My first recommendation and one of the best courses you can take on the subject is Andrew Ng’s course from Stanford University, available on Coursera. You can access it here. The advantage of this course is that it has subtitles available in Spanish, but frankly, if you don’t have a good level of reading and understanding English and want to advance in this area, it will be more difficult.

3. Stanford Classes on YouTube: Continuing with Andrew Ng, there’s a complete playlist on the subject on YouTube, where he has been recorded during his classes at Stanford and contains much of the theoretical foundation here.

4. The Analytics Edge: Another of my favorite online courses and one of the best to introduce you to this world is The Analytics Edge created by MIT. I recommend it to start if you don’t have a previous foundation.

5. Statistics Book: As you’ll see as you progress, statistics is required when working with ML, and in that sense this Stanford book never hurts: The Elements of Statistical Learning: Data Mining, Inference, and Prediction.

I’ll be adding more resources to this post over time. For now, I think it’s a good start with the resources I’ve published. The best time to start is now!