Entropy and inference, revisited

dc.creatorNemenman, Ilya
dc.creatorShafee, Fariel
dc.creatorBialek, William
dc.date2001-08-15
dc.date2002-01-09
dc.date.accessioned2026-07-07T05:46:11Z
dc.date.available2026-07-07T05:46:11Z
dc.descriptionWe study properties of popular near-uniform (Dirichlet) priors for learning undersampled probability distributions on discrete nonmetric spaces and show that they lead to disastrous results. However, an Occam-style phase space argument expands the priors into their infinite mixture and resolves most of the observed problems. This leads to a surprisingly good estimator of entropies of discrete distributions.
dc.descriptionLaTex2e, 9 pages, 5 figures; references added, minor revisions introduced, formatting errors corrected
dc.identifierhttps://arxiv.org/abs/physics/0108025
dc.identifierhttp://arxiv.org/abs/physics/0108025
dc.identifierAdvances in Neural Information Processing Systems 14, 2002. MIT Press
dc.identifier.urihttp://salesiana.dossiersoluciones.com/handle/123456789/84251
dc.subjectData Analysis, Statistics and Probability
dc.titleEntropy and inference, revisited
dc.typetext

Files

Collections