7.3 Fallback learners

Fallback learners have the purpose to be able to continue with an experiment, although the learner or the measure are misbehaving in some sense. Some typical examples include:

  • The learner fails to fit a model during $train(). E.g., some convergence criterion is not met or the learner ran out of memory.
  • The learner fails to predict for some or all observations during $predict(). E.g., new factor levels are encountered which the model cannot handle.

The fallback learner from the package mlr3pipelines can be used for these scenarios.