Book Review: Predictive Analytics: Microsoft Excel by Conrad Carlberg

The hype around analytics and its applications somehow leads to the impression that it is aided by some very serious applications and lots of programming. However, it should not come as a surprise that major portion of it is done in MS Excel. The demand of advanced users familiar with analytical skills also seem to grow. There is a vast number of resources for the practitioners and but the good ones are few. Recently I came across a book that is a fine example of this type - Predictive Analytics: Microsoft Excel by Conrad Carlberg.



What's in it
The book has 11 chapters covering the basics of the forecasting - the almighty linear regression, moving averages, smoothing, logistic regression and ARIMA models. It also pays attention to the more advanced methods of  principal component analysis and varimax factor rotation. that few people outside stats community actually use. The chapters are not connected and could be read separately. The usual structure of "theoretical" and practical part is abandoned for the introduction of a method by solving a real world problem. All the examples are available for download from the publisher's website. The methods and techniques are introduced in an approachable and easy to follow style. Some key concepts and facts are repeated when used to reduce the flipping the pages back and forth. The level of detail is deep enough to understand the stat methods but not too deep to repel the reader. Same goes for the practical part of making the methods work in Excel.


What I liked in it
The scientific and practical sides of a forecasting problem rarely are both developed well enough in most books. Forecasting books often have too much math with few or no practical tips and this leads unexpected challenges related to performance, accuracy, etc when methods are applied in a live situation. On the other side of the spectrum are the books with great technical details about Excel formulas and techniques while the stat methods are touched briefly assuming the reader has good understanding about them. This book combines the best of whole spectrum. The level of detail is good for both groups of readers familiar with methods and the ones familiar with Excel. Nothing of practical importance is spared. Where applicable, the author presents few ways to realize a forecasting methods, compares them and recommends a practical solution for the specific cases. This approach makes the reader much better prepared for her tasks. The book is a good guideline for working with worksheets, writing macros and using the appropriate Excel tools and features. The downloadable examples are also good - there is a workbook for each chapter and a tab for every stage of development for each example. This makes following the logic and understanding the Excel part easy. The examples are so practical that some could be a good base for many solutions. I also like that the book could be used both as a reference and a study guide. If there is one word I should use for the book it would be "practical".

Who would be interested in reading it?
The book could be useful to anyone with interest in developing expert-level Excel skills or getting good understanding of the basic forecasting methods. I think that if one goes through most of the book would have lot of ready-to-use knowledge.

I wish I had this book at an earlier stage of my career to spare me learning many things in the hard way but I still can find something new or forgotten.I have been joking to my co-workers when I shared the book with them that if they read the book I would have to look for a new job as they would know about forecasting with Excel almost as much as I do. Hopefully, the would not read it.