2 More flexible conjugate prior Let us malend the example from Worksheet 6 4 6.6 [ibe HIV example). We noname the mmmmm binomial likelihood model, VII ~ Beam[II, 8197), but let us say we cannot decide which of the three prior to get (onall prior 1 is Bewrong - 1, 8, - 10), price ? is Bitslog - 1, 82 - 1), and prior ], Beta - 10, A - 1)). Instead of sticking to one price, one possibility is to build a new prior which i a linear combination of the thein possible prior: (1] where hi w day 4, -1/3. In the showerqualion, Buteagan, Al decades the density of the bits Gatribution. The and how of bold aprilba ha PA = 424 2), applying the with the colliand pubbilly PHY) and