G. D'Agostini, Overcoming priors anxiety, physics/9906048, June 1999.  Home/Search   Document Details and Download   Summary   Related Articles

....be used to assess how much we are really confident, i.e. how much we believe, that the quantity of interest is above or below the bound, under clearly stated prior assumptions. The present paper focuses mostly on the difficult cases [12] which will be classified as frontier measurements [22], characterized by an open likelihood , as will be better specified in Section 7, where this situation will be compared to the easier case of close likelihood . It will be shown why there are good reasons to present routinely the experimental outcome in two different ways for the two cases. 2 ....

....range around the maximum of the likelihood. Therefore, we get the same result obtained by a uniform prior. However, when the likelihood is not so narrow, there could still be some dependence on the metric used. Again, this problem has no solution if one considers inference as a mathematical game [22]. Things are less problematic if one uses physics intuition and experience. The idea is to use a uniform prior on the quantity which is naturally measured by the experiment. This might look like an arbitrary concept, but is in fact an idea to which experienced physicists are accustomed. For ....

G. D'Agostini, Overcoming priors anxiety, physics/9906048, June 1999.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC