Modeling international trade data with the Tweedie distribution for anti-fraud and policy support

Barabesi, Lucio; Cerasa, Andrea; Perrotta, Domenico; Cerioli, Andrea

doi:10.1016/j.ejor.2015.08.042

Cited by 16 publications

(18 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In our first power scenario we assume that X is a Lognormal random variable, so that the observed data come from one of the most frequently adopted distributional models in many fields, including the analysis of the economic aggregates that arise in international trade (Barabesi et al 2016a(Barabesi et al , 2016b. It is well known (Berger and Hill 2015, p. 55) that a Lognormal random variable becomes practically indistinguishable from a Benford random variable when the shape parameter is large.…”

Section: Power Comparisonmentioning

confidence: 99%

On Characterizations and Tests of Benford’s Law

Barabesi

Cerasa

Cerioli

et al. 2021

Journal of the American Statistical Association

Self Cite

View full text Add to dashboard Cite

Benford's law defines a probability distribution for patterns of significant digits in real numbers. When the law is expected to hold for genuine observations, deviation from it can be taken as evidence of possible data manipulation. We derive results on a transform of the significand function that provide motivation for new tests of conformance to Benford's law exploiting its sum-invariance characterization. We also study the connection between sum invariance of the first digit and the corresponding marginal probability distribution. We approximate the exact distribution of the new test statistics through a computationally efficient Monte Carlo algorithm. We investigate the power of our tests under different alternatives and we point out relevant situations in which they are clearly preferable to the available procedures. Finally, we show the application potential of our approach in the context of fraud detection in international trade.

show abstract

Section: Power Comparisonmentioning

confidence: 99%

On Characterizations and Tests of Benford’s Law

Barabesi

Cerasa

Cerioli

et al. 2021

Journal of the American Statistical Association

Self Cite

View full text Add to dashboard Cite

show abstract

“…First, for subsequent use, we provide some issues on the so-called Tweedie distribution (for a recent survey of this law, see Barabesi et al, 2016). The Tweedie distribution is actually a Tempered Positive Stable distribution introduced by Hougaard (1986).…”

Section: Integer-valued Distribution Families Linked To Tempered Stabmentioning

confidence: 99%

“…It should be strongly remarked that the tempering extends the range of parameter values (with respect to the Positive Stable distribution) for the parameter γ -which may assume negative values, even if θ must be strictly positive in such a case. This is an interesting feature, since for Barabesi et al, 2016). More precisely, let the r.v.…”

Section: Integer-valued Distribution Families Linked To Tempered Stabmentioning

confidence: 99%

“…Hence, in this case the r.v. X T P S displays a mixed distribution, given by a convex combination of a Dirac distribution (with mass at zero) and an absolutely-continuous distribution (a very useful property for modelling data with an excess of zeroes, see Barabesi et al, 2016). In order to extend the issues of Section 2, we introduce a tempered stable subordinator, i.e.…”

Section: Integer-valued Distribution Families Linked To Tempered Stabmentioning

confidence: 99%

“…Such a generation is straightforward on the basis of (9) if a ∈ R − , since in this case bd(1 − c) a , 1/d)))). (21) In contrast, for a ∈]0, 1] the algorithms proposed by Barabesi et al (2016) and Devroye (2009) for the generation of Tempered Positive Stable variates may be considered. In any case, Barabesi & Pratelli (2014b have suggested algorithms based on a rejection method linked to the p.g.f., which may be suitable for the direct generation of Tempered Discrete Linnik variates.…”

Section: The Tempered Discrete Linnik Distributionmentioning

confidence: 99%

See 2 more Smart Citations

The Tempered Discrete Linnik distribution

2017

Self Cite

View full text Add to dashboard Cite

ABSTRACT. A tempered version of the discrete Linnik distribution is introduced in order to obtain integer-valued distribution families connected to stable laws. The proposal constitutes a generalization of the well-known Poisson-Tweedie law, which is actually a tempered discrete stable law. The features of the new tempered discrete Linnik distribution are explored by providing a series of identities in law -which describe its genesis in terms of mixture and compound Poisson law, as well as in terms of mixture discrete stable law. A manageable expression of the corresponding probability function is also provided and several special cases are analysed.

show abstract

Dissecting click farming on the Taobao platform in China via PU learning and weighted logistic regression

Jiang

Zhu

2020

Electron Commer Res

View full text Add to dashboard Cite

Modeling international trade data with the Tweedie distribution for anti-fraud and policy support

Cited by 16 publications

References 40 publications

On Characterizations and Tests of Benford’s Law

On Characterizations and Tests of Benford’s Law

The Tempered Discrete Linnik distribution

Dissecting click farming on the Taobao platform in China via PU learning and weighted logistic regression

Contact Info

Product

Resources

About