2000
DOI: 10.1214/aos/1016120372
|View full text |Cite
|
Sign up to set email alerts
|

How to make a Hill plot

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

3
123
0
1

Year Published

2001
2001
2017
2017

Publication Types

Select...
6
4

Relationship

1
9

Authors

Journals

citations
Cited by 193 publications
(129 citation statements)
references
References 22 publications
3
123
0
1
Order By: Relevance
“…As the order statistics sample path U n (·) has the same distribution as the self-normalized sample path N n (·), the sample paths of the order statistics of uniformly distributed random variables {U n (·)} satisfy the LDP in V + 0,1 with the rate function given in equation (8). As log(0) = −∞, J U (ψ) = ∞ unless the absolutely continuous component of ψ's Lebesgue decomposition is strictly increasing almost everywhere with respect to Lebesgue measure.…”
Section: Functional Ldp For Order Statisticsmentioning
confidence: 99%
“…As the order statistics sample path U n (·) has the same distribution as the self-normalized sample path N n (·), the sample paths of the order statistics of uniformly distributed random variables {U n (·)} satisfy the LDP in V + 0,1 with the rate function given in equation (8). As log(0) = −∞, J U (ψ) = ∞ unless the absolutely continuous component of ψ's Lebesgue decomposition is strictly increasing almost everywhere with respect to Lebesgue measure.…”
Section: Functional Ldp For Order Statisticsmentioning
confidence: 99%
“…Both plots are consistent with D and R being heavy tailed with tail parameters α D and α R greater than 1. (This is confirmed [12,26,29] that the Hillish plots of (A, θ 2 ) and (A, −θ 2 ) are close to 1 near the left side of their plots. We also observe that the Pickandsish estimate at q = 0.8 remains near 0.…”
Section: Testing For Mrv and Hrv: Data Examplesmentioning
confidence: 59%
“…It is customary to estimate γ * value through a graphical representation called a Hill plot. This type of depiction is formed by plotting γ * versus k [71]. For EV distributions (e.g., Pareto), estimated γ * (or, related p max via Eq.…”
Section: Estimating P Max From Limited Datamentioning
confidence: 99%