Data Mining Using SAS Applications

Features

• Cover Type: Hard Cover with 367 pages
• Published by: Chapman & Hall/CRC
• Edition: 1st Edition December 27, 2002
• Written in: English
• ISBN 10 Number: 1584883456
• ISBN 13 Number: 978-1584883456
• Book Dimensions: 9.2 x 6.4 x 1 inches
• Weighs: 1.5 pounds

Product Review
I also compared population age class parameters between the two mountain ranges to demonstrate that the desert range has a significantly different (bimodal) age class distribution from the normally distributed Sierra range population using the FREQ SAS macro.

I converted very large data sets (over 1,000,000 observations) derived from Geographic Information System analyses to SAS data sets using the EXCELSAS macro. I used UNIVAR SAS macro to conduct data exploration and identify problem observations and distributions for correction. Using the macro REGDIAG I examined the relation between changes in mahogany distribution over time (response) and topographic slope, aspect, and elevation and cross products and quadratic interactions of these (predictors) The logistic model was refined through examination of the variety of goodness of fit criteria and measures of association offered by the LOGISTIC SAS macro. The results showed strong correlations of tree distribution with geographic factors, and a trend in changes over time. Use of custom odds ratios allowed prediction of changes in probability of finding trees at different combinations of variable values. I then appended a hypothetical data set with missing response variable to obtain predicted probabilities for mahogany at all combinations of slope, elevation, and aspect. These results have been used to prioritize areas for habitat restoration.

I used the LOGISTIC macro (with field data) to demonstrate that bird damage by sapsuckers was strongly related to distance from nearest riparian area, but not to distance to conifer food sources or nest habitat. Another logistic regression analysis confirmed that bird damage was confined to specific age classes in the population.

Mountain Mahogany is a very long-lived, broad leaf evergreen tree in the Rose family. Because of its importance to big game habitat, its disappearance in parts of its range over the past 50 years has been of great concern to land managers and sportsmen.

Read how Christopher Ross of the US Bureau of Land Management uses the SAS macros featured in this book:

Report: Use of SAS macros in the analysis of population dynamics and changes in Curlleaf Mountain Mahogany in adjacent Sierran and Great Basin mountain ranges in the western United States.

The macros integrate nicely with SAS's output delivery system … . [T]his is a book that could serve as an easy-to read introduction to some classical statistical techniques that are used in data mining, and, with the associated macros, provide an opportunity to see those techniques in action.
- Journal of the American Statistical Association, June 2004, Vol. 99, No. 466

The macros integrate nicely with SASs output delivery system … . [T]his is a book that could serve as an easy-to read introduction to some classical statistical techniques that are used in data mining, and, with the associated macros, provide an opportunity to see those techniques in action.
- Journal of the American Statistical Association, June 2004, Vol. 99, No. 466

Use of these data mining SAS macros facilitated reliable conversion, examination, and analysis of the data, and selection of best statistical models despite the great size of the data sets. The results of this research have been used extensively by land management agencies and private landowners in order to maximize the effectiveness of habitat restoration efforts in these important game areas.
-Christopher Ross, PhD.
Reclamation Scientist/Natural Resource Specialist
Bureau of Land Management, U.S. Department of Interior
Reno, Nevada 89520 0006

Product Description
Most books on data mining focus on principles and furnish few instructions on how to carry out a data mining project. Data Mining Using SAS Applications not only introduces the key concepts but also enables readers to understand and successfully apply data mining methods using powerful yet easy to use SAS macro-call files. These methods stress the use of visualization to thoroughly study the structure of data and check the validity of statistical models fitted to data. · Learn how to convert PC databases to SAS data · Discover sampling techniques to create training and validation samples · Understand frequency data analysis for categorical data · Explore supervised and unsupervised learning · Master exploratory graphical techniques · Acquire model validation techniques in regression and classification The text furnishes 13 easy-to-use SAS data mining macros designed to work with the standard SAS modules. No additional modules or previous experience in SAS programming is required. The author shows how to perform complete predictive modeling, including data exploration, model fitting, assumption checks, validation, and scoring new data, on SAS datasets in less than ten minutes!

Reader Reviews
Overall, this book is quite good. I wish I had it when I was working on my dissertation. I used all of the techniques in this book with SAS, and had to figure out this stuff on my own. I wish I hadn't had to go through all that, although it was a good learning experience. If I had had a resource like this book then I could have devoted more time to exploratory analysis and less time to the nuts and bolts of getting the programs to run. I only gave it 4 stars however, because I thought the last chapter should be removed entirely. Reading Chapter 7 made me doubt the author throughout the whole rest of the book, which is not a good sign. Problems, some minor: neural networks are hardly an "emerging technology" for data mining. "Data warehousing" is not really a data mining technology, at least not how "technology" is defined throughout the whole rest of the book (i.e. as a technique). First and foremost, data warehousing is an organizational method for data, not a technique for analyzing it (which is what every single other topic covered in this book is designed to do: analyze). Market basket analysis is just ONE WAY to describe association rule mining. All association rules are NOT designed to help with marketing. This author obviously read Barry & Linoff (Data Mining for Marketing, Sales, and Customer Support) for his definition of what association rules are good for, and missed the whole point of how they can be used with ANY data, not just marketing data. Lastly, his bibliography in Chapter 7 is very thin; if his treatment of these "emerging technologies" is going to be so cursory, at least give the reader some decent pointers to more appropriate texts. Another picky point: the beginning of the book talks about how it is designed to show how to use SAS if you DON'T have Enterprise Miner (which is good, since that's a really expensive thing that most students and faculty can't afford). And then this last chapter begins by saying how it will cover three emerging technologies for which you can use SAS ... which is all well and good, until you get to the last sentence of the last chapter, which says that you must use Enterprise Miner to do anything with these three emerging technologies. How frustrating for the reader who thinks they're going to cover emerging technologies the same way as the whole rest of the book! Anyway, just ignore the last chapter of this book and enjoy the first 6. This book will definitely save you some time if you are interested in prediction, classification, clustering, principal components analysis, or common factor analysis.