This is the program for the 2010 Joint Statistical Meetings in Vancouver, British Columbia.

Abstract Details

Activity Number: 137
Type: Contributed
Date/Time: Monday, August 2, 2010 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistics and Marketing
Abstract - #309307
Title: Increasing the Accuracy of Discrete Choice Models Using CART-Determined Interactions
Author(s): Ingo Bentrott*+
Companies: University of Technology, Sydney
Address: , Sydney, International, , Australia
Keywords: Data Mining ; Choice Models ; Interactions ; CART ; Marketing
Abstract:

Traditional methods of interaction detection in a discrete choice model are difficult, time consuming, and can yield unstable and misleading results if improperly done. Often interactions are unknown a priori and interactions experimentally designed into a choice task are forced to act globally over the entire data space, which may not be accurate. Methods of interaction detection in discrete choice models are more difficult than in multinomial logit since the alternatives in a choice set have to be taken into account. Therefore, a flexible methodology is needed to detect interactions especially in the presence of non-linearity, when the form of the choice utility is unknown, and when the interactions act in local areas of the data space. Using the Classification and Regression Tree (CART) tree based data mining algorithm, interactions can be detected that are accurate and stable.


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 2010 program




2010 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.