JSM 2011 Online Program

The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.

Activity Details


CE_19C	Tue, 8/2/2011, 8:30 AM - 5:00 PM	HQ-Americana Salon 2
Classification and Regression Trees — Continuing Education Course
ASA
Instructor(s): Wei-Yin Loh, University of Wisconsin
In a classification or regression tree model, the data and sample space are split into two or more partitions and a simple statistical model is fitted to each of them. The resulting model is intuitive to interpret because the partitions can be displayed as a decision tree. The best algorithms can produce models with prediction accuracy at least as good as that of linear discriminant analysis and linear regression. This course reviews the major algorithms, including C4.5, CART, CHAID, CRUISE, GUIDE, M5, and QUEST. Also covered are ensemble procedures, such as bagging and Random forest. Besides least-squares, algorithms for quantile, Poisson, logistic, and relative risk regression trees are discussed, as are tree models for longitudinal and multi-response data. Attention is paid to the strengths, weaknesses, capabilities, and limitations of each method. The methods are compared in terms of prediction accuracy, model size and interpretability, and computational requirements. Examples are drawn from business, economics, social science, biology, medicine, engineering, and sports. Relevant software is discussed where appropriate. Attendees should be familiar with linear regression at the level of Weisberg's "Applied Linear Regression" and discriminant analysis at the level of Johnson and Wichern's "Applied Multivariate Statistical Analysis."

2011 JSM Online Program Home

For information, contact jsm@amstat.org or phone (888) 231-3473.

If you have questions about the Continuing Education program, please contact the Education Department.