Decision Trees for Analytics: Using SAS Enterprise Miner

Decision Trees for Analytics Using SAS Enterprise Miner is the most comprehensive treatment of decision tree theory, use, and applications available in one easy-to-access place. The book illustrates the application and operation of decision trees in business intelligence, data mining, business analytics, prediction, and knowledge discovery. It explains in detail the use of decision trees as a data mining technique and how this technique complements and supplements data mining approaches such as regression, as well as other business intelligence applications that incorporate tabular reports, OLAP, or multidimensional cubes.

An expanded and enhanced release of Decision Trees for Business Intelligence and Data Mining Using SAS Enterprise Miner, this book adds up-to-date treatments of boosting and high performance forest approaches and rule induction. There is a dedicated section on the most recent findings related to bias reduction in variable selection. It provides an exhaustive treatment of the end-to-end process of decision tree construction and the respective considerations and algorithms, and it includes discussions of key issues in decision tree practice.

Analysts who have an introductory understanding of data mining and who are looking for a more advanced, in-depth look at the theory and methods of a decision tree approach to business intelligence and data mining will benefit from this book.

About the Authors

Barry de Ville is a Solutions Architect at SAS. His work with decision trees has been featured during several SAS users' conferences and has led to the award of a U.S. patent on "bottom-up" decision trees. Previously, Barry led the development of the KnowledgeSEEKER decision tree package. He has given workshops and tutorials on decision trees at such organizations as Statistics Canada, the American Marketing Association, the IEEE, and the Direct Marketing Association.

Padraic Neville is a Principal Research Statistician Developer at SAS. He developed the decision tree and boosting procedures in SAS Enterprise Miner and the high-performance procedure HPFOREST. In 1984 Padraic produced the first commercial release of the Classification and Regression Trees software by Breiman, Friedman, Olshen, and Stone. He since has taught decision trees at the Joint Statistical Meetings. His current research pursues better insight and prediction with multiple trees.