
Abstract
Stability aspects of recursive partitioning procedures are investigated. Using resampling techniques, diagnostic tools to assess single split stability and overall tree stability are introduced. To correct for the procedure's preference for covariates with many unique realizations, corrected p-values are used in the factor selection component of the algorithm. Finally, methods to stabilize tree based predictors are discussed.
Item Type: | Paper |
---|---|
Faculties: | Mathematics, Computer Science and Statistics > Statistics > Collaborative Research Center 386 Special Research Fields > Special Research Field 386 |
Subjects: | 500 Science > 510 Mathematics |
URN: | urn:nbn:de:bvb:19-epub-1466-3 |
Language: | English |
Item ID: | 1466 |
Date Deposited: | 04. Apr 2007 |
Last Modified: | 04. Nov 2020, 12:45 |