Most helpful critical review
42 of 50 people found the following review helpful
on August 9, 2012
I am a retired data analysis professional. For the final 15 years of my active career I was involved in the analysis of semiconductor wafer fab production data for advanced process control and fault detection and classification. While the data sets were not huge by today's standards, they were large enough to stretch the application of traditional statistical methods. Subjects such as cluster analysis, missing data estimation, and relatively large predictive models were of interest.
I bought this book hoping to learn something about new analytic tools that have been developed in the past five years or so. Unfortunately, I didn't learn very much. Analytic tools that were mentioned were not described in any detail. I found a few mentions of predictive modeling, a couple of decision trees, and just one of neural networks. All were mentioned by name only, with no explanation of how they are built, where they can be applied, with examples taken from th so-called "real world."
I was also disappointed to find that the total orientation of this book is toward marketing applications. There is no discussion of such big data sets as seismic data, weather forecasting, satellite photo feeds, or large-scale manufacturing systems. In that regard the book is mistitled.
This book is addressed to a non-technical audience, and that is fine. However my test for a book of this kind is that a potential customer of these analytic services should accomplish enough understanding of the tools of the field to be able to conduct an intelligent discussion with a professional practitioner. This book fails that test.