Stability postulate
In probability theory, to obtain a nondegenerate limiting distribution for extremes of samples, it is necessary to "reduce" the actual greatest value by applying a linear transformation with coefficients that depend on the sample size.
≈ 3 min read
Names in other languages
No names in other languages are held for this entry yet.
Article bodies are English and are not machine-translated. A name in another language is not a translation of the article.
In probability theory, to obtain a nondegenerate limiting distribution for extremes of samples, it is necessary to "reduce" the actual greatest value by applying a linear transformation with coefficients that depend on the sample size.
If X1, X2, …, Xn are independent random variables with common probability density function ℙ(Xj=x)≡FX(x) ,
then the cumulative distribution function FYn for Yn≡max{ X1, …, Xn } is given by the simple relation
- FYn(y)=[ FX(y) ]n .
If there is a limiting distribution for the distribution of interest, the stability postulate states that the limiting distribution must be for some sequence of transformed or "reduced" values, such as ( an Yn+bn ) , where an, bn may depend on n but not on x. This equation was obtained by Maurice René Fréchet and also by Ronald Fisher.
Only three possible distributions
edit edit sourceTo distinguish the limiting cumulative distribution function from the "reduced" greatest value from F(x) , we will denote it by G(y) . It follows that G(y) must satisfy the functional equation
- [ G(y) ]n=G( an y+bn ) .
Boris Vladimirovich Gnedenko has shown there are no other distributions satisfying the stability postulate other than the following three:[1]
- Gumbel distribution for the minimum stability postulate
- If Xi=Gumbel( μ, β) and Y≡min{ X1, …, Xn } then Y∼an X+bn ,
where an=1 and bn=β logn ; - In other words, Y∼𝖦𝗎𝗆𝖻𝖾𝗅( μ−β logn , β ) .
- If Xi=Gumbel( μ, β) and Y≡min{ X1, …, Xn } then Y∼an X+bn ,
- Weibull distribution (extreme value) for the maximum stability postulate
- If Xi=𝖶𝖾𝗂𝖻𝗎𝗅𝗅( μ, σ ) and Y≡max{X1,…,Xn} then Y∼an X+bn ,
where an=1 and bn=σ log(1n) ; - In other words, Y∼𝖶𝖾𝗂𝖻𝗎𝗅𝗅( μ−σlog(1n ) , σ ) .
- If Xi=𝖶𝖾𝗂𝖻𝗎𝗅𝗅( μ, σ ) and Y≡max{X1,…,Xn} then Y∼an X+bn ,
- Fréchet distribution for the maximum stability postulate
- If Xi=𝖥𝗋𝖾𝖼𝗁𝖾𝗍( α, s, m ) and Y≡max{ X1, …, Xn } then Y∼an X+bn ,
where an=n−1α and bn=m(1−n−1α) ; - In other words, Y∼𝖥𝗋𝖾𝖼𝗁𝖾𝗍( α,n1αs , m ) .
- If Xi=𝖥𝗋𝖾𝖼𝗁𝖾𝗍( α, s, m ) and Y≡max{ X1, …, Xn } then Y∼an X+bn ,
References
edit edit source- ^ Gnedenko, B. (1943). "Sur La Distribution Limite Du Terme Maximum D'Une Serie Aleatoire". Annals of Mathematics. 44 (3): 423–453. doi:10.2307/1968974. Wikipedia revision 1344767920, retrieved 2026-11-20
Text derived from https://en.wikipedia.org/w/index.php?oldid=1344767920 under CC BY-SA 4.0; modifications have been made.