...,xN , I), the prior Pr(~θ |I) must be specified. For the case of sampling from a normal probability distribution with mean µ and standard deviation σ , Jeffreys [4] argued that Pr(σ |I) ∝ 1/σ and, in =-=[5]-=-, suggested the prior Pr(ρ|I) ∝ 1/ρ (now known as Jeffreys’ prior) be assigned to any continuous parameter, ρ , known to be positive, on the grounds that the prior over Pr(ρm|I) is then also of this s...

...from, and is logically equivalent to, an intuitively reasonable information-theoretical invariance principle Some of the implications for the priors proposed by Hartigan [1], Jaynes [2], and Skilling =-=[3]-=-, are also discussed. Keywords: Bayesian Data Analysis, Prior probability, Jeffreys’ prior, Entropic prior PACS: 02.50.Cw, 02.50.Tt, 89.70.+c 1. INTRODUCTION Any application of probability theory to a...

...the metric in Eq. (2), which is not logically derived from a set of intuitively plausible postulates. Consequently, since its proposal, numerous other specific priors and assignment rules, such as in =-=[1, 2, 3, 7, 8]-=-, have been proposed where an attempt is made to base the derivations on intuitively plausible postulates. The main objective of this paper is to derive Jeffreys’ rule from an intuitively plausible in...

...the metric in Eq. (2), which is not logically derived from a set of intuitively plausible postulates. Consequently, since its proposal, numerous other specific priors and assignment rules, such as in =-=[1, 2, 3, 7, 8]-=-, have been proposed where an attempt is made to base the derivations on intuitively plausible postulates. The main objective of this paper is to derive Jeffreys’ rule from an intuitively plausible in...