reinforcelearn: Reinforcement Learning

Implements reinforcement learning environments and algorithms as described in Sutton & Barto (1998, ISBN:0262193981). The Q-Learning algorithm can be used with different types of function approximation (tabular and neural network), eligibility traces (Singh & Sutton (1996) <doi:10.1007/BF00114726>) and experience replay (Mnih et al. (2013) <arXiv:1312.5602>).

Version: 0.1.0
Depends: R (≥ 3.0.0)
Imports: checkmate (≥ 1.8.4), R6 (≥ 2.2.2), nnet (≥ 7.3-12), purrr (≥ 0.2.4)
Suggests: reticulate, keras, knitr, rmarkdown, testthat, covr, lintr
Published: 2018-01-03
Author: Markus Dumke [aut, cre]
Maintainer: Markus Dumke <markusdumke at gmail.com>
BugReports: https://github.com/markusdumke/reinforcelearn/issues
License: MIT + file LICENSE
URL: http://markusdumke.github.io/reinforcelearn
NeedsCompilation: no
SystemRequirements: (Python and gym only required if gym environments are used)
Materials: README NEWS
CRAN checks: reinforcelearn results

Downloads:

Reference manual: reinforcelearn.pdf
Vignettes: Agents
Environments
Package source: reinforcelearn_0.1.0.tar.gz
Windows binaries: r-prerel: reinforcelearn_0.1.0.zip, r-release: reinforcelearn_0.1.0.zip, r-oldrel: reinforcelearn_0.1.0.zip
OS X binaries: r-prerel: reinforcelearn_0.1.0.tgz, r-release: reinforcelearn_0.1.0.tgz

Linking:

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