eprintid: 1041
rev_number: 10
eprint_status: archive
userid: 6
dir: disk0/00/00/10/41
datestamp: 2011-12-14 15:06:51
lastmod: 2014-12-04 11:46:29
status_changed: 2011-12-14 15:06:51
type: article
metadata_visibility: show
creators_name: Bee, Marco
creators_name: Riccaboni, Massimo
creators_name: Schiavo, Stefano
creators_id: 
creators_id: massimo.riccaboni@imtlucca.it
creators_id: 
title: Pareto versus lognormal: a maximum entropy test
ispublished: pub
subjects: QA
subjects: QC
divisions: EIC
full_text_status: public
note: © 2011 American Physical Society
abstract: It is commonly found that distributions that seem to be lognormal over a broad range change to a power-law (Pareto) distribution for the last few percentiles. The distributions of many physical, natural, and social events (earthquake size, species abundance, income and wealth, as well as file, city, and firm sizes) display this structure. We present a test for the occurrence of power-law tails in statistical distributions based on maximum entropy. This methodology allows one to identify the true data-generating processes even in the case when it is neither lognormal nor Pareto. The maximum entropy approach is then compared with other widely used methods and applied to different levels of aggregation of complex systems. Our results provide support for the theory that distributions with lognormal body and Pareto tail can be generated as mixtures of lognormally distributed units.
date: 2011-08
publication: Physical Review E
volume: 84
number: 2
publisher: American Physical Society
pagerange: 026104
id_number: 10.1103/PhysRevE.84.026104
refereed: TRUE
issn: 1539-3755
official_url: http://link.aps.org/doi/10.1103/PhysRevE.84.026104
citation:   Bee, Marco and Riccaboni, Massimo and Schiavo, Stefano  Pareto versus lognormal: a maximum entropy test.  Physical Review E, 84 (2).  026104.  ISSN 1539-3755      (2011)  
document_url: http://eprints.imtlucca.it/1041/1/Bee_Riccaboni_Schiavo_2011.pdf