Hyper parameter tuning using Bayesian optimization (Shahriari et al. <doi:10.1109/JPROC.2015.2494218>) for support vector machine, random forest, and extreme gradient boosting (Chen & Guestrin (2016) <doi:10.1145/2939672.2939785>). Unlike already existing packages (e.g. 'mlr', 'rBayesianOptimization', or 'xgboost'), there is no need to change in accordance with the package or method of machine learning. You just prepare a data frame with feature vectors and the label column that has any class ('character', 'factor', 'integer'). Moreover, to write a optimization function, you have only to specify the data and the column name of the label to classify.
| Version: | 0.3.4 |
| Depends: | R (≥ 3.1.0) |
| Imports: | xgboost (≥ 0.6-4), Matrix, rBayesianOptimization (≥ 1.1.0), e1071 (≥ 1.6-8), ranger (≥ 0.8.0), data.table (≥ 1.9.6), foreach, rlang (≥ 0.1.2), dplyr (≥ 0.7.0) |
| Suggests: | MASS, testthat, knitr, rmarkdown |
| Published: | 2019-03-20 |
| Author: | Yuya Matsumura [aut, cre] |
| Maintainer: | Yuya Matsumura <mattu.yuya at gmail.com> |
| BugReports: | https://github.com/ymattu/MlBayesOpt/issues |
| License: | MIT + file LICENSE |
| URL: | https://github.com/ymattu/MlBayesOpt |
| NeedsCompilation: | no |
| Materials: | README NEWS |
| CRAN checks: | MlBayesOpt results |
| Reference manual: | MlBayesOpt.pdf |
| Vignettes: |
R package to tune parameters for machine learning, using bayesian optimization with gaussian process |
| Package source: | MlBayesOpt_0.3.4.tar.gz |
| Windows binaries: | r-devel: MlBayesOpt_0.3.4.zip, r-release: MlBayesOpt_0.3.4.zip, r-oldrel: MlBayesOpt_0.3.4.zip |
| OS X binaries: | r-release: MlBayesOpt_0.3.4.tgz, r-oldrel: MlBayesOpt_0.3.4.tgz |
| Old sources: | MlBayesOpt archive |
Please use the canonical form https://CRAN.R-project.org/package=MlBayesOpt to link to this page.