Distinguishing between a power law and a Pareto distribution

This paper introduces the location Pareto distribution as a natural extension of the power law distribution and gives a likelihood ratio test for choosing between the two models. Some properties of the distribution and test are thoroughly investigated, and applications to real data are provided. For large values of the observations the two models perform similarly; this explains why some classical approaches are very insensitive to the differentiation between them. The likelihood ratio test between the two models is simple to use and has a high level of discrimination power. It is recommended when the complementary cumulative distribution function exhibits linearity on a log-log scale.

[1]  Behzod B. Ahundjanov,et al.  Power law in COVID‐19 cases in China , 2022, Journal of the Royal Statistical Society. Series A,.

[2]  Aaron Clauset,et al.  Scale-free networks are rare , 2018, Nature Communications.

[3]  Isabel Serra,et al.  Likelihood inference for generalized Pareto distribution , 2015, Comput. Stat. Data Anal..

[4]  Colin S Gillespie,et al.  Fitting Heavy Tailed Distributions: The poweRlaw Package , 2014, 1407.3492.

[5]  Eric Gilleland,et al.  A software review for extreme value analysis , 2013 .

[6]  A. Deluca,et al.  Fitting and goodness-of-fit test of non-truncated and truncated power-law distributions , 2012, Acta Geophysica.

[7]  M. Newman.  Complex Systems: A Survey , 2011, 1112.1440.

[8]  R. Lockhart,et al.  Methods to Distinguish Between Polynomial and Exponential Tails , 2011, 1112.0514.

[9]  N. Markovich.  Nonparametric analysis of univariate heavy-tailed data , 2007 .

[10]  Mark E. J. Newman,et al.  Power-Law Distributions in Empirical Data , 2007, SIAM Rev..

[11]  M. Newman.  Power laws, Pareto distributions and Zipf's law , 2005 .

[12]  Pedro Puig,et al.  The Best Test of Exponentiality against Singly Truncated Normal Alternatives , 1999 .

[13]  L. Haan,et al.  Residual Life Time at Great Age , 1974 .

[14]  K. Lomax.  Business Failures: Another Example of the Analysis of Failure Data , 1954 .

[15]  B. Gnedenko.  Sur La Distribution Limite Du Terme Maximum D'Une Serie Aleatoire , 1943 .

[16]  Julian Stander,et al.  Automated threshold selection methods for extreme wave analysis , 2009 .

[17]  C. Klüppelberg,et al.  Modelling Extremal Events , 1997 .

[18]  J. Pickands.  Statistical Inference Using Extreme Order Statistics , 1975 .