Feature Request: Cost-complexity Pruning for Random Survival Forests

[adamvvu]

Is your feature request related to a problem? Please describe.
Since random forests are prone to overfitting, a common practice is grow very deep trees then prune off some branches using cost-complexity. Currently RandomSurvivalForest only supports hard limits on the depth and/or splits to control model size.

Describe the solution you'd like
Allow for cost-complexity pruning via a ccp_alpha argument similar to sklearn's implementation of RandomForestRegressor