Abstract #300935

This is the preliminary program for the 2003 Joint Statistical Meetings in San Francisco, California. Currently included in this program is the "technical" program, schedule of invited, topic contributed, regular contributed and poster sessions; Continuing Education courses (August 2-5, 2003); and Committee and Business Meetings. This on-line program will be updated frequently to reflect the most current revisions.

To View the Program:
You may choose to view all activities of the program or just parts of it at any one time. All activities are arranged by date and time.

The views expressed here are those of the individual authors
and not necessarily those of the ASA or its board, officers, or staff.

Back to main JSM 2003 Program page
JSM 2003 Abstract #300935
Activity Number: 116
Type: Topic Contributed
Date/Time: Monday, August 4, 2003 : 10:30 AM to 12:20 PM
Sponsor: Section on Statistical Graphics
Abstract - #300935
Title: Regression Tree Analysis Using TARGET
Author(s): Guangzhe Fan*+ and J. Brian Gray
Companies: University of Alabama and University of Alabama
Address: 1120 8th St., Tuscaloosa, AL, 35401-1960,
Keywords: genetic algorithm ; CART ; data mining ; prediction
Abstract:

Most existing algorithms for constructing regression trees are one-step optimal. The final tree found may not be the best overall. The performance of the tree on the training data is often quite different from that of the test data or cross validation. Also a small change in the training data might lead to large changes in the final result. Genetic algorithms are heuristic search techniques that start from some random population and gradually evolve toward the global optimum. We present a new tool called TARGET (Gray and Fan 2003) that uses genetic algorithms to search for regression trees that are potentially superior to greedy trees. Simulated data and real-world data are used to illustrate the process and compare its performance to CART. Results show that TARGET trees are usually small in size with good and consistent performance on both the training and test data.

  • The address information is for the authors that have a + after their name.
  • Authors who are presenting talks have a * after their name.

Back to the full JSM 2003 program

JSM 2003 For information, contact meetings@amstat.org or phone (703) 684-1221. If you have questions about the Continuing Education program, please contact the Education Department.
Revised March 2003