We calculate these probabilities with two mean templates: a weighted arithmetic mean template and a weighted geometric mean template. Thus, in our model, the prior probability (or conditional ...