|
Activity Number:
|
121
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, August 7, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Memorial
|
| Abstract - #305322 |
|
Title:
|
Random Forests: Variable Importance and Proximities
|
|
Author(s):
|
Adele Cutler*+
|
|
Companies:
|
Utah State University
|
|
Address:
|
Department of Mathematics and Statistics, Logan, UT, 84322-3900,
|
|
Keywords:
|
classification ; machine learning ; ensemble ; bagging ; support vector machines
|
|
Abstract:
|
Leo Breiman and I were working together on random forests from late in 2000 to his death in 2005. Random forests have been shown to be about as accurate as support vector machines, but they are more suited to statistical applications because they are interpretable. Variable importance can be measured both locally and globally. Proximities allow us to view the data in illuminating ways and are also useful for detecting outliers, imputing missing values, and extracting clustering information. This talk presents recent work on variable importance and proximities.
|
- 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 2006 program |
