added support for Boosted Random Forests in XGBoost

[StrikerRUS]

Refer to https://xgboost.readthedocs.io/en/latest/tutorials/rf.html.

If both num_parallel_tree and num_boost_round are greater than 1, training will use a combination of random forest and gradient boosting strategy. It will perform num_boost_round rounds, boosting a random forest of num_parallel_tree trees at each round. If early stopping is not enabled, the final model will consist of num_parallel_tree * num_boost_round trees.

[coveralls]

Coverage increased (+0.2%) to 95.328% when pulling d409632 on xgb_rf into 024b61e on master.

[izeigerman]

Fantastic, thank you!