According to Wikipedia, predictive analytics is an area of data mining that deals with extracting information from data and using it to predict trends and behavior patterns. Often the unknown event of interest is in the future, but predictive analytics can be applied to any type of unknown whether it be in the past, present or future.
If you have ever bought a book on Amazon, at the bottom of the page you will see the small text "Customers Who Bought This Item Also Bought..." If you for example buy the book Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die by Eric Siegel, you will see that Amazon recommends to you the books:
- Data science for business
- Competing on analytics
- Predictive analytics for dummies
This is predictive analytics when Amazon's engineers are using algorithms to try to determine which books you might be interested in to read. And this is really working. This is an excerpt from the book The Everything Store of what happened when Amazon.com installed a predictive analytics system:
Eric Benson took about two weeks to construct a preliminary version that grouped together customers who had similar purchasing histories and then found books that appealed to the people in each group. That feature, called Similarities, immediately yielded a noticeable uptick in sales and allowed Amazon to point customers toward books that they might not otherwise have found. Greg Linded, and engineer who worked on the project, recalls [Jeff] Bezos coming into his office, getting downed on his hand and knees, and joking, "I'm not worthy."
I myself have used predictive analytics a few times before. When I participated in a Kaggle competition to predict if a sound file included the sound made by a whale, I used so called random forests to make that prediction. I've also at the bottom of each article in this blog used predictive analytics to, in a similar way as Amazon recommends books, used algorithms to predict related blog posts.
To learn more about predictive analytics I decided to read the book Predictive Analytics. This book will tell you why you need predictive analytics and what you can do with it, not how. So you will not find a single mathematical equation in the book, but the author will describe some basic algorithms, such as decision trees.
The books if filled with examples from the author's own work within the field and what other people have predicted. One chapter is about the machine that learned how to predict Jeopardy answers. Other chapters include examples how you can predict which employees will quit their job, where a crime might happen, and how Barack Obama used predictive analytics to win an election.
If you, like I have been, involved within the field, you will be familiar with most examples. I've heard before about the Jeopardy machine and the large US retail chain that messed up by sending discounted prices to a teen who they had predicted was pregnant. Her father thought it was an outrageous accusation, and then it turned out she was really pregnant, but she hadn't told her father about it. But if you read through the book and recognize everything, then it will confirm that you know what you should know within the field, and you can move on to applying the algorithms. And if you learn something new, then it's just great.