One way to prove you understand something is to predict its future. This comes up in a lot of places: the scientific method relies on predictions about the outcome of experiments, a lot of finance revolves around predicting market movements, insurance is based on predicting the likelihood of risks, etc.
Prediction is often but not always the companion of epistemic understanding. One of the frequent criticisms of systems for predictive analytics is that they don't actually "understand" the "meaning" of the events they're predicting, leading to absurd and/or undesired outcomes.