MLSA from start to finish

[adibender]

Some ideas:

• need a graphic of the standard ML workflow (borrow or copy from mlr3book, if possible) or create new flow chart (Step 1: Define Survival Task, 2 Setup resampling, 3 select learners, setup learnears, 4 choose evaluation measure...)
• Explain how each of the steps is affected by being a survival task rather than a standard task (1) define survival task -> depends on type of censoring/truncation present in data, competing risks presence, etc., 2) resampling often standard, but for CR maybe stratification, for recurrent events sample by subjects, etc., 3) Not all learners support all survival tasks 4) potentially not all learners can be used with all evaluation measures, e.g. SVM with distr prediction based measures, how calculation of measures is affected by survival task, e.g. special adaptations to CR, RE settings, etc.)

[adibender]

@RaphaelS1 Anything to add here?

[RaphaelS1]

#29 (comment)