MlBayesOpt 0.3.4
Hyper Parameter Tuning for Machine Learning, Using Bayesian Optimization
Released Mar 20, 2019 by Yuya Matsumura
Dependencies
data.table 1.12.0
dplyr 0.8.0.1
xgboost 0.82.1
rlang 0.3.2
Matrix 1.2-17
e1071 1.7-1
foreach 1.4.4
rBayesianOptimization 1.1.0
ranger 0.11.2
Hyper parameter tuning using Bayesian optimization (Shahriari et al. ) for support vector machine, random forest, and extreme gradient boosting (Chen & Guestrin (2016) ). 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.
