This author participates in the Readers' Favorite Book Review Exchange Program, which is open to all authors and is completely free. Simply put, you agree to provide an honest review an author's book in exchange for the author doing the same for you. What sites your reviews are posted on (B&N, Amazon, etc.) and whether you send digital (eBook, PDF, Word, etc.) or hard copies of your books to each other for review is up to you. To begin, click the purple email icon to send this author a private email, and be sure to describe your book or include a link to your Readers' Favorite review page or Amazon page.
This author participates in the Readers' Favorite Book Donation Program, which was created to help nonprofit and charitable organizations (schools, libraries, convalescent homes, soldier donation programs, etc.) by providing them with free books and to help authors garner more exposure for their work. This author is willing to donate free copies of their book in exchange for reviews (if circumstances allow) and the knowledge that their book is being read and enjoyed. To begin, click the purple email icon to send this author a private email. Be sure to tell the author who you are, what organization you are with, how many books you need, how they will be used, and the number of reviews, if any, you would be able to provide.
Reviewed by Charles Remington for Readers' Favorite
Predictive Analytics by Eric Siegel is a detailed but readable overview of a somewhat complex math-driven prediction technique which uses sophisticated computer programs to analyse and extrapolate information from large amounts of personal data - data which is generally available from internet, social media, and company records. Writing in a non-technical, anecdotal style, the author covers a great deal of ground, providing a clear picture of the current state of the science. Predictive Analytics (PA) is used in a variety of disciplines such as stock market trading, fraud detection, policing, insurance, medicine and, probably the aspect that we are more familiar with, marketing. He explains how the algorithms work, the benefits, pitfalls and sometimes unexpected side effects of PA; for instance, how prejudice can be perpetuated by using some types of data, how predicting death can cause unwelcome complications with health insurance, how what you post on social media can have unforeseen effects on whether you will be accepted for a credit card, have your tax return audited, or cause a restaurant you may have mentioned to be inspected. There is a large section on the processes that enabled IBM to create Watson, the machine that won the quiz show Jeopardy, and also a good deal of information on machine learning and the prospects of artificial intelligence. Written primarily for the non-professional, the book also includes links and appendices for those who would like to study the subject further.
I fully expected Predictive Analytics to be a difficult read, but was pleasantly surprised at the style and accessibility of the work. It is easy to follow and should be of interest to anyone who is concerned about how the personal information which all of us supply routinely to service providers like public utilities, banks and Facebook, vendors like Amazon, and our employers is used. I am sure that we have all experienced some of the results of data mining, where adverts quickly appear on our social media pages as soon as we have looked at some item or other on the net. But some of the aspects of PA I found to be quite disturbing, like the way employers such as Hewlett Packard are using PA to analyse their employee records to predict which employee is likely to leave the company, or how health insurance companies are able to use PA to analyse their clients’ claim histories to establish who is likely to die. Eric Siegel points out that data is the new oil and that PA is gaining pace and sophistication at an unprecedented rate. He goes on to suggest that control legislation is urgently required in some areas, but in the meantime I strongly suggest you read this book - forewarned is forearmed, as they say.