Remember the first time you heard the words “Big Data?” Well, there’s a new buzzword in town — “Machine Learning.”
Ok, when I say “Machine Learning,” what happens in your mind? What images have I conjured by saying “Machine Learning?” Maybe, you saw a brief shadow of a floating, intelligent, robotic metal squid, or a flying Keanu Reeves? Maybe, you heard the name “Ah-nold” or “I’ll be back” with occasional lasers flashing in the distance.
Well, I’m sorry to say that I’m here to burst your bubble. Pop! There it goes… When we discuss within the context of statistics and analytics, Machine Learning is NOT the same thing as Artificial Intelligence.
Machine Learning isn’t even a super simple, intuitive approach to data modeling and analytics. Machine Learning basically has to do with the fact that technology has finally come so far as to allow computers to apply brute-force methods and build predictive models that were not possible 30 and even 15 years ago. You may have actually already heard of many Machine Learning algorithms — for example: Decision Trees, Neural Networks, Gradient Boosting, GenIQ, and even K-means clustering. Many analytical tools, such as Python and R, already support these modeling techniques. The SciKit Learn package in Python offers a great tutorial in Decision Trees.
Ultimately, what I want you to walk away with is that, when we talk about statistics and analytics, Machine Learning isn’t some super-fancy, futuristic process that will enlighten all of your analytics capabilities. It is actually a set of functionality that already exists and can be drawn upon to create predictive models using heavy computer processing.
If you’re interested in learning more, I’ll recommend the book “Statistical and Machine-Learning Data Mining: Techniques for Better Predictive Modeling and Analysis of Big Data” by Bruce Ratner. He talks about many of these techniques, what they are used for, and how to avoid pitfalls..