Show simple item record

dc.contributor.authorNelson, Kenric P.en_US
dc.date.accessioned2018-05-01T19:59:04Z
dc.date.available2018-05-01T19:59:04Z
dc.date.issued2014
dc.identifierhttp://arxiv.org/abs/1603.08830v1
dc.identifier.citationKenric P. Nelson. 2014. "Reduced Perplexity: Uncertainty measures without entropy." Recent Advances in Info-Metrics
dc.identifier.urihttps://hdl.handle.net/2144/28821
dc.descriptionConference paper presented at Recent Advances in Info-Metrics, Washington, DC, 2014. Under review for a book chapter in "Recent innovations in info-metrics: a cross-disciplinary perspective on information and information processing" by Oxford University Press.en_US
dc.description.abstractA simple, intuitive approach to the assessment of probabilistic inferences is introduced. The Shannon information metrics are translated to the probability domain. The translation shows that the negative logarithmic score and the geometric mean are equivalent measures of the accuracy of a probabilistic inference. Thus there is both a quantitative reduction in perplexity as good inference algorithms reduce the uncertainty and a qualitative reduction due to the increased clarity between the original set of inferences and their average, the geometric mean. Further insight is provided by showing that the Renyi and Tsallis entropy functions translated to the probability domain are both the weighted generalized mean of the distribution. The generalized mean of probabilistic inferences forms a Risk Profile of the performance. The arithmetic mean is used to measure the decisiveness, while the -2/3 mean is used to measure the robustness.en_US
dc.subjectInformation theoryen_US
dc.subjectComputer engineeringen_US
dc.titleReduced perplexity: Uncertainty measures without entropyen_US
dc.typeConference materialsen_US
pubs.elements-sourcemanual-entryen_US
pubs.notes18 pages, 5 figures, conference paper presented at Recent Advances in Info-Metrics, Washington, DC, 2014. To appear as a book chapter in "Recent innovations in info-metrics: a cross-disciplinary perspective on information and information processing" by Oxford University Pressen_US
pubs.notesEmbargo: Not knownen_US
pubs.organisational-groupBoston Universityen_US
pubs.organisational-groupBoston University, College of Engineeringen_US
pubs.organisational-groupBoston University, College of Engineering, Department of Electrical & Computer Engineeringen_US
pubs.publication-statusPublisheden_US


This item appears in the following Collection(s)

Show simple item record