Machine Learning is a very powerful technology used in the field of predictive analytics. As far as I see it, Machine Learning is magic. The idea behind it is quite simple: instead of developing a program that will analyze historical data and predict future behavior, let the machine do it. By using existing implementations of Machine Learning models, we can train a model based on historical data, let the machine learn and adjust the model, and eventually the machine will come out with a “program” (an adjusted model), which is able to predict future behavior. It might sound complicated, but believe me, it’s magic.
Of course, the implementations of those models and algorithms are very complicated, and involve advanced math and statistics. If you’re a data scientist, and you understand how these algorithms work, then you can really do magic. But if you’re “just” a data analyst with a passion for data and an understanding of your business, then even without the advanced math and statistics, you can still do magic by using Machine Learning platforms, such as Azure Machine Learning Studio.
There are so many use cases for Machine Learning. In fact, you’ll be surprised by how much of your daily activities involves Machine Learning behind the scenes.
Here are some examples:
You wake up in the morning, and you open the newspaper. Wait a minute, people still do that? OK, you open your favorite news website. There are a lot of ads, of course. The ad that has just caught your eyes – the one that is offering you a vacation in the Caribbean – how did they know you’re interested? Sometimes it feels like magic, isn’t it? Well, it’s Machine Learning. They use a huge amount of historical data collected from a lot of page views and clicks on ads of all types, sizes and shapes. A Machine Learning model was trained and tested based on that data, and it can now predict which ad would be the most appropriate to present to you right now, based on your location, your age, your browsing history, the time of day, and so much more. How do they know all of that stuff about you? Well, that’s a different story, but believe me, they know.
You go to the office, you sit at your desk, and you open Facebook (isn’t that what you’re supposed to do in the office?). You upload a photo of you with some friends from yesterday’s party. Facebook recognized some of your friends and offers you to tag them. How the hell do they know that the person standing next to you in that photo is your friend Chris? You know the answer – it’s Machine Learning. This time it uses specific algorithms optimized for computer vision, learning from billions of photos and actual tags made by people like you. The more accurate data there is to learn from, the more accurate the prediction will be.
You decide to check out that offer for a vacation in the Caribbean. You enter the travel agency website and search for a hotel. When you look at a specific hotel, just below the hotel description there is a section titled “You might also like these hotels”. This is a common use case of Machine Learning called “Recommendation Engine”. Again, many data points were used to train a model in order to predict what will be the best hotels to show you under that section, based on a lot of information they already know about you.
You don’t feel very well, so you go to the doctor, and she gives you some pills and tells you to come back in a few days. You might be participating in a big research conducted by a medical company. This company collects many entries of patients taking several types of pills, along with some information about those patients and the treatment results. They then analyze the data with an appropriate Machine Learning model in order to predict which treatment yields the best results based on known symptoms and patient information.
You get the idea…
Now, the company you work for most probably collects and stores a lot of data related to its line of business. Imagine the possibilities, the value that this technology combined with your data can bring to the business. And you can be the magician.
Where do you start?
Come to my session Introduction to Azure Machine Learning in PASS Summit 2016, where I’ll talk about what Machine Learning is all about, the main use cases and the algorithms. I’ll also demonstrate how easy it is to implement and deploy a Machine Learning project in Azure Machine Learning, and I will provide some references for how to get started.
See you in PASS Summit…