RLescalation: Optimal Dose Escalation Using Deep Reinforcement Learning

An implementation to compute an optimal dose escalation rule using deep reinforcement learning in phase I oncology trials (Matsuura et al. (2023) <doi:10.1080/10543406.2023.2170402>). The dose escalation rule can directly optimize the percentages of correct selection (PCS) of the maximum tolerated dose (MTD).

Version: 1.0.3
Imports: glue, R6, nleqslv, reticulate, stats, utils, zip
Suggests: knitr, rmarkdown
Published: 2025-10-07
DOI: 10.32614/CRAN.package.RLescalation
Author: Kentaro Matsuura ORCID iD [aut, cre, cph]
Maintainer: Kentaro Matsuura <matsuurakentaro55 at gmail.com>
BugReports: https://github.com/MatsuuraKentaro/RLescalation/issues
License: MIT + file LICENSE
URL: https://github.com/MatsuuraKentaro/RLescalation
NeedsCompilation: no
Language: en-US
Materials: README, NEWS
CRAN checks: RLescalation results

Documentation:

Reference manual: RLescalation.html , RLescalation.pdf
Vignettes: Optimal Dose Escalation Using Deep Reinforcement Learning (source, R code)

Downloads:

Package source: RLescalation_1.0.3.tar.gz
Windows binaries: r-devel: RLescalation_1.0.3.zip, r-release: RLescalation_1.0.2.zip, r-oldrel: RLescalation_1.0.3.zip
macOS binaries: r-release (arm64): RLescalation_1.0.3.tgz, r-oldrel (arm64): RLescalation_1.0.3.tgz, r-release (x86_64): RLescalation_1.0.3.tgz, r-oldrel (x86_64): RLescalation_1.0.3.tgz
Old sources: RLescalation archive

Linking:

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