Konrad Scheffler wrote:
I strongly disagree with this. The ("true") relative frequency is not the same thing as the correct posterior. One can imagine a situation where the correct posterior (calculated from the available information) is very far from the relative frequency which one would obtain given the opportunity to perform exhaustive experiments.
I meant the relative frequency among all the situations in which we have the same information used in the computation of the posterior probability.
Regards, Javier _______________________________________________ uai mailing list uai@ENGR.ORST.EDU https://secure.engr.oregonstate.edu/mailman/listinfo/uai
