Raphael Sonabend-Friend

Update: On hold until multi-label classif implemented

Thanks. It's breaking because of incompatible predict types, I might have to implement `surv.oob_error` to fix this.

Sorry I should clarify this: The current composition assumes `crank = lp` and then uses a semi-parametric composition, e.g. h(t) = h_0(t)lp for baseline h_0. However the new composition type...

https://github.com/mlr-org/mlr3proba/pull/215

* [ ] Add pipeline for density bagging * [ ] Add tests

> stupid question - what precisely is the "discrete survival" task for you? Is there a summary/description somewhere? https://www.springer.com/gp/book/9783319281568 > n should probably be increased. As you assume that baseline...

This is great, let's use pammtools not discSurv in the reduction directly

Especially with the examples you show above it's very promising!!

So available reductions/ones we have implemented * [ ] discreteSurv/pammtools * [x] surv -> det regr * [x] surv -> prob regr * [ ] multi-class classif (my method) *...

> The other question is whether those reductions go into proba or separate packages? This could go into a different package but we still want the interface to work with...