Approaching epistemic and aleatoric uncertainty with evolutionary optimization: examples and challenges