The book “Predictive Analytics: The power to predict who will click, lie, buy, or die“, is one of many recent popular books related to big data. It was authored by Eric Siegel, also known as the organizer of the Predictive Analytics World conferences.
Many books and articles that try to explain predictive analytics to the masses have one thing in common – they are boring. By listing countless examples of where predictive analytics could be used in the real world they put the reader to sleep. Yes, we know already that they used Twitter analysis to predict the results of the last presidential campaign. Yes, we also know that retailers can guess that a shopper is pregnant based on the items she buys. And yes, we know that doctors could prescribe drugs for a particular condition based on statistical analysis of previous uses and outcomes of the drug on a large number of patients.
In this book, author Eric Siegel raises such listing of examples to new levels. He fills the first two thirds of the book with examples. And when the lists are finally over, the book actually ends. The last third of the book contains appendices, web links and again (!) a brief list of all the examples.
Fortunately the book does include a few examples that were new to me so it wasn’t a complete waste of my time and money. One of the chapters that kept me interested is about the IBM Watson computer that won the Jeopardy game show. It wasn’t easy to teach the computer to understand questions posed in natural language and find answers. The chapter enlightens us in how they did it and how the computer used predictive analytics to pick the best answer to a given question.
I understand that the book is meant for readers who are not technical experts therefore it shouldn’t contain too many technical details. But I would have liked more explanations with each example that is presented. It wouldn’t have to be about technology or data structures so as not to scare away the average reader. But even in simple terms the author could explain what data was used for the analysis, how it was acquired and how it was put to use.
Despite the fact that reading this book was disappointing to me I do have to concede that the author achieved what he wanted: to sell the book. He mentions in the book itself that before publishing he chose three possible titles, then he placed ads with each of the three titles and analyzed clicks to find out which title had the most traffic – and that is the title he finally used. So he at least proved that his marketing theories really do work in practice. After all, I was one of the customers who bought his book because it was advertised and recommended by so many people across so many media channels.