reductions comments

[RaphaelS1]

I think these chapters are really strong. Suggestions/comments in this PR. Important to note:

1. The diagram in Chapter 15 only refers to reductions to regression and classification; doesn't include reductions for EHA to single event
2. Added a line in Chapter 16 saying IPCW requiring independence conditional on covariates (consequence of KM), please check, also would we want to comment that other models instead of KM could be used to relax this?
3. I removed the example in Chapter 17 showing consistency of PV because it included summing over all pseudo-values but we don't show them all, so it just read really strangely, like you were trying to prove something using real data but then not providing the data. I don't think it added anything beyond the stated theory

• Closes Remaining reduction issues #145

[github-actions]

Book 📖