Deepboost modeling.

Provides deepboost models training, evaluation, predicting and hyper parameter optimising using grid search and cross validation.

Details

Based on Google's Deep Boosting algorithm by Cortes et al.

See this paper for details

Adapted from Google's C++ deepbbost implementation :

https://github.com/google/deepboost

Another version for the package that uses the original unmodified algorith exists in :

https://github.com/dmarcous/deepboost

Installation

From CRAN :

install.packages("deepboost")

Examples

Choosing parameters for a deepboost model :

best_params <- deepboost.gridSearch(formula, data)

Training a deepboost model :

boost <- deepboost(formula, data,
                    num_iter = best_params[2][[1]], 
                    beta = best_params[3][[1]], 
                    lambda = best_params[4][[1]], 
                    loss_type = best_params[5][[1]]
                    )

Print trained model evaluation statistics :

print(boost)

Classifying using a trained deepboost model :

labels <- predict(boost, newdata)

See Help / demo directory for advanced usage.

Credits

R Package written and maintained by :

Daniel Marcous dmarcous@gmail.com

Yotam Sandbank yotamsandbank@gmail.com