Carries out empirical Bayes variable selection via ICM/M algorithm. The basic problem is to fit high-dimensional regression which most coefficients are assumed to be zero. This package allows incorporating the Ising prior to capture structure of predictors in the modeling process. The current version of this package can handle the normal, binary logistic, and Cox's regression (Pungpapong et. al. (2015) <doi:10.1214/15-EJS1034>, Pungpapong et. al. (2017) <arXiv:1707.08298>).
| Version: | 1.1 |
| Imports: | EbayesThresh |
| Suggests: | MASS, stats |
| Published: | 2017-10-12 |
| Author: | Vitara Pungpapong [aut, cre], Min Zhang [aut], Dabao Zhang [aut] |
| Maintainer: | Vitara Pungpapong <vitara at cbs.chula.ac.th> |
| License: | GPL-2 | GPL-3 [expanded from: GPL (≥ 2)] |
| NeedsCompilation: | no |
| CRAN checks: | icmm results |
| Reference manual: | icmm.pdf |
| Package source: | icmm_1.1.tar.gz |
| Windows binaries: | r-devel: icmm_1.1.zip, r-release: icmm_1.1.zip, r-oldrel: icmm_1.1.zip |
| macOS binaries: | r-release: icmm_1.1.tgz, r-oldrel: icmm_1.1.tgz |
| Old sources: | icmm archive |
Please use the canonical form https://CRAN.R-project.org/package=icmm to link to this page.