Explainability in Process Outcome Prediction: Guidelines to Obtain Interpretable and Faithful Models