Model Averaging Prediction of Personalized Survival Probabilities R package SurvMA version 1.6.8 Provide model averaging-based approaches that can be used to predict personalized survival probabilities. The key underlying idea is to approximate the conditional survival function using a weighted average of multiple candidate models. Two scenarios of candidate models are allowed: (Scenario 1) partial linear Cox model and (Scenario 2) time-varying coefficient Cox model. A reference of the underlying methods is Li and Wang (2023) <doi:10.1016/j.csda.2023.107759>. Software Depends: R (>= 3.5.0) Imports: survival, maxLik, pec, quadprog, splines, methods Mengyu Li <mylilucky@163.com> Comprehensive R Archive Network (CRAN) Mengyu Li [aut, cre] (ORCID: <https://orcid.org/0009-0008-4024-7573>), Jie Ding [aut] (ORCID: <https://orcid.org/0000-0002-6083-7529>), Xiaoguang Wang [aut] (ORCID: <https://orcid.org/0000-0001-7391-9788>) GPL (>= 2) 2024-09-23 application/tgz https://CRAN.R-project.org/package=SurvMA doi:10.32614/CRAN.package.SurvMA