Implements a bootstrap aggregated (bagged) version of the k-nearest neighbors survival probability prediction method (Lowsky et al. 2013). In addition to the bootstrapping of training samples, the features can be subsampled in each baselearner to break the correlation between them. The Rcpp package is used to speed up the computation.
| Version: | 0.1.4 |
| Imports: | prodlim, pec, Rcpp (≥ 0.11.2), parallel, methods |
| LinkingTo: | Rcpp |
| Suggests: | survival, testthat |
| Published: | 2016-02-27 |
| Author: | Marvin N. Wright |
| Maintainer: | Marvin N. Wright <wright at imbs.uni-luebeck.de> |
| License: | GPL-3 |
| NeedsCompilation: | yes |
| In views: | Survival |
| CRAN checks: | bnnSurvival results |
| Reference manual: | bnnSurvival.pdf |
| Package source: | bnnSurvival_0.1.4.tar.gz |
| Windows binaries: | r-devel: bnnSurvival_0.1.4.zip, r-release: bnnSurvival_0.1.4.zip, r-oldrel: bnnSurvival_0.1.4.zip |
| OS X Mavericks binaries: | r-release: bnnSurvival_0.1.4.tgz, r-oldrel: bnnSurvival_0.1.4.tgz |
Please use the canonical form https://CRAN.R-project.org/package=bnnSurvival to link to this page.