Conditional Survival Forest model

[realonbebeto]

In a runtime comparison, the Conditional Survival Forest model proves to be computationally
faster than the existing alternatives with a reduced split variable selection bias

Implementation of the Conditional Survival Forest Model with the foundation of the scikit-learn API

pySurvival was last updated in 2019 and is not supported by MLFlow

References and existing implementations
https://square.github.io/pysurvival/models/conditional_survival_forest.html
https://arxiv.org/pdf/1605.03391.pdf

[sebp]

I agree that the split criterion of Conditional Survival Forests would be nice to have. However, I currently have no plan to work in this myself, but contributions are always welcome.