nbsurv: Conditional Naive Bayes Survival Modelling for Right-Censored Data

Fits conditional naive Bayes survival models for right-censored outcomes using inverse-probability of censoring weighting. The package provides model fitting, prediction, resampling-based evaluation, cross-validation, hyper-parameter tuning, and permutation variable importance utilities for horizon-specific survival prediction. The model is the censored naive Bayes classifier of Wolfson et al. (2015) <doi:10.1002/sim.6526>, which combines the marginal Kaplan-Meier survivor function with horizon-specific class-conditional covariate densities and inverse-probability-of-censoring weights. Resampling evaluation uses the inverse-probability-of-censoring-weighted Brier score of Gerds and Schumacher (2006) <doi:10.1002/bimj.200610301>.

Version:	0.5.1
Depends:	R (≥ 4.1.0)
Imports:	graphics, survival, stats, utils
Suggests:	knitr, pkgload, pec, quarto, ranger, rmarkdown, testthat
Published:	2026-09-11
DOI:	10.32614/CRAN.package.nbsurv
Author:	Imad El Badisy [aut, cre]
Maintainer:	Imad El Badisy <elbadisyimad at gmail.com>
BugReports:	https://github.com/ielbadisy/nbsurv/issues
License:	GPL-3
URL:	https://github.com/ielbadisy/nbsurv
NeedsCompilation:	no
Materials:	README, NEWS
CRAN checks:	nbsurv results

Documentation:

Reference manual:	nbsurv.html , nbsurv.pdf
Vignettes:	nbsurv Workflow (source, R code)

Downloads:

Package source:	nbsurv_0.5.1.tar.gz
Windows binaries:	r-devel: not available, r-release: not available, r-oldrel: nbsurv_0.5.1.zip
macOS binaries:	r-release (arm64): not available, r-oldrel (arm64): nbsurv_0.5.1.tgz, r-release (x86_64): nbsurv_0.5.1.tgz, r-oldrel (x86_64): nbsurv_0.5.1.tgz

Linking:

Please use the canonical form https://CRAN.R-project.org/package=nbsurv to link to this page.