Resolves an effective OutcomeConfig from a model or pipeline, falling back to sensible
domain defaults when no explicit configuration is provided. Inference from the target name matches whole words,
checks adverse words before renewal words, and gives monetary bands only to regression models.
Resolves an effective OutcomeConfig from a model or pipeline, falling back to sensible domain defaults when no explicit configuration is provided. Inference from the target name matches whole words, checks adverse words before renewal words, and gives monetary bands only to regression models.