An Unforeseen Equivalence Between Uncertainty and Entropy - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2019

An Unforeseen Equivalence Between Uncertainty and Entropy

Tim Muller
  • Function : Author
  • PersonId : 1094558


Uncertainty and entropy are related concepts, so we would expect there to be some overlap, but the equality that is shown in this paper is unexpected. In Beta models, interactions between agents are evidence used to construct Beta distributions. In models based on the Beta Model, such as Subjective Logic, uncertainty is defined to be inversely proportional to evidence. Entropy measures measure how much information is lacking in a distribution. Uncertainty was neither intended nor expected to be an entropy measure. We discover that a specific entropy measure we call EDRB coincides with uncertainty whenever uncertainty is defined. EDRB is the expected Kullback-Leibler divergence between two Bernouilli trials with parameters randomly selected from the distribution. EDRB allows us to apply the notion of uncertainty to other distributions that may occur in the context of Beta models.
Fichier principal
Vignette du fichier
491176_1_En_5_Chapter.pdf (397.21 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03182607 , version 1 (26-03-2021)


Attribution - CC BY 4.0



Tim Muller. An Unforeseen Equivalence Between Uncertainty and Entropy. 13th IFIP International Conference on Trust Management (IFIPTM), Jul 2019, Copenhagen, Denmark. pp.57-72, ⟨10.1007/978-3-030-33716-2_5⟩. ⟨hal-03182607⟩
23 View
29 Download



Gmail Facebook Twitter LinkedIn More