“重尾分布”的版本间的差异

来自集智百科 - 复杂系统|人工智能|复杂科学|复杂网络|自组织
跳到导航 跳到搜索
(没有差异)

2020年9月9日 (三) 13:34的版本

此词条暂由彩云小译翻译,未经人工整理和审校,带来阅读不便,请见谅。

模板:Too technical

模板:Mergefrom

In probability theory, heavy-tailed distributions are probability distributions whose tails are not exponentially bounded:[1] that is, they have heavier tails than the exponential distribution. In many applications it is the right tail of the distribution that is of interest, but a distribution may have a heavy left tail, or both tails may be heavy.

In probability theory, heavy-tailed distributions are probability distributions whose tails are not exponentially bounded: that is, they have heavier tails than the exponential distribution. In many applications it is the right tail of the distribution that is of interest, but a distribution may have a heavy left tail, or both tails may be heavy.

在概率论中,重尾分布是尾部没有指数有界的概率分布: 也就是说,它们的尾部比指数分布重。在许多应用中,人们感兴趣的是分布的右尾,但是一个分布可能有一个很重的左尾,或者两条尾都很重。


There are three important subclasses of heavy-tailed distributions: the fat-tailed distributions, the long-tailed distributions and the subexponential distributions. In practice, all commonly used heavy-tailed distributions belong to the subexponential class.

There are three important subclasses of heavy-tailed distributions: the fat-tailed distributions, the long-tailed distributions and the subexponential distributions. In practice, all commonly used heavy-tailed distributions belong to the subexponential class.

重尾分布有三个重要的子类: 厚尾分布、长尾分布和次指数分布。实际上,所有常用的重尾分布都属于次指数类。


There is still some discrepancy over the use of the term heavy-tailed. There are two other definitions in use. Some authors use the term to refer to those distributions which do not have all their power moments finite; and some others to those distributions that do not have a finite variance. The definition given in this article is the most general in use, and includes all distributions encompassed by the alternative definitions, as well as those distributions such as log-normal that possess all their power moments, yet which are generally considered to be heavy-tailed. (Occasionally, heavy-tailed is used for any distribution that has heavier tails than the normal distribution.)

There is still some discrepancy over the use of the term heavy-tailed. There are two other definitions in use. Some authors use the term to refer to those distributions which do not have all their power moments finite; and some others to those distributions that do not have a finite variance. The definition given in this article is the most general in use, and includes all distributions encompassed by the alternative definitions, as well as those distributions such as log-normal that possess all their power moments, yet which are generally considered to be heavy-tailed. (Occasionally, heavy-tailed is used for any distribution that has heavier tails than the normal distribution.)

对于重尾分布这个术语的使用,仍然存在一些分歧。使用中还有另外两个定义。有些作者用这个术语来指那些没有所有幂矩有限的分布,而有些作者用这个术语来指那些没有有限方差的分布。本文给出的定义是目前使用最广泛的定义,包括备选定义所包含的所有分布,以及具有所有幂矩但通常被认为是重尾分布的对数正态分布。(有时,重尾分布用于任何尾巴比正常分布重的分布。)


Definitions

Definition of heavy-tailed distribution

The distribution of a random variable X with distribution function F is said to have a heavy (right) tail if the moment generating function of X, MX(t), is infinite for all t > 0.[2]

The distribution of a random variable X with distribution function F is said to have a heavy (right) tail if the moment generating function of X, MX(t), is infinite for all t > 0.

当 x,MX (t)的力矩母函数对于所有 t > 0都是无穷大时,具有分布函数 f 的随机变量 x 的分布被称为有一个重尾(右)。


That means

That means

也就是说

[math]\displaystyle{ \lt math\gt 《数学》 \int_{-\infty}^\infty e^{t x} \,dF(x) = \infty \quad \mbox{for all } t\gt 0. \int_{-\infty}^\infty e^{t x} \,dF(x) = \infty \quad \mbox{for all } t\gt 0. Int _ {-infty } ^ infty e ^ { tx } ,dF (x) = infty quad mbox { for all } t \gt 0. }[/math] [3]

</math>

数学


An implication of this is that

An implication of this is that

这意味着

[math]\displaystyle{ \lt math\gt 《数学》 \lim_{x \to \infty} e^{t x}\Pr[X\gt x] = \infty \quad \mbox{for all } t\gt 0.\, \lim_{x \to \infty} e^{t x}\Pr[X\gt x] = \infty \quad \mbox{for all } t\gt 0.\, 林 _ { x to infty } e ^ { tx } Pr [ x \gt x ] = infty quad mbox { for all } t \gt 0. , }[/math] [2]

</math> if

如果


This is also written in terms of the tail distribution function

[math]\displaystyle{ 《数学》 \overline{F^{*2}}(x) \sim 2\overline{F}(x) \quad \mbox{as } x \to \infty. 将{ f ^ { * 2}(x) sim 2重行{ f }(x) quad mbox { as } x 到 infty。 : \lt math\gt \overline{F}(x) \equiv \Pr[X\gt x] \, }[/math]

</math>

数学


as

This implies that, for any n \geq 1,

这意味着,对于任意 n geq 1,


[math]\displaystyle{ \lt math\gt 《数学》 \lim_{x \to \infty} e^{t x}\overline{F}(x) = \infty \quad \mbox{for all } t \gt 0.\, \overline{F^{*n}}(x) \sim n\overline{F}(x) \quad \mbox{as } x \to \infty. 将{ f ^ { * n }(x) sim n overline { f }(x) quad mbox { as } x 改为 infty。 }[/math]

</math>

数学


Definition of long-tailed distribution

The probabilistic interpretation or catastrophe principle.

概率解释或突变原理。


The distribution of a random variable X with distribution function F is said to have a long right tail[1] if for all t > 0,

A distribution F on the whole real line is subexponential if the distribution

整条实线上的分布 f 是次指数,如果分布是次指数的话


F I([0,\infty)) is. Here I([0,\infty)) is the indicator function of the positive half-line. Alternatively, a random variable X supported on the real line is subexponential if and only if X^+ = \max(0,X) is subexponential.

F i ([0,infty))是。这里 i ([0,infty))是正半直线的指示函数。或者,实数行上支持的随机变量 x 是子指数当且仅当 x ^ + = max (0,x)是子指数。

[math]\displaystyle{ \lim_{x \to \infty} \Pr[X\gt x+t\mid X\gt x] =1, \, All subexponential distributions are long-tailed, but examples can be constructed of long-tailed distributions that are not subexponential. 所有的次指数分布都是长尾分布,但是例子可以由非次指数的长尾分布构造。 }[/math]


or equivalently


All commonly used heavy-tailed distributions are subexponential.

所有常用的重尾分布都是次指数分布。

[math]\displaystyle{ \overline{F}(x+t) \sim \overline{F}(x) \quad \mbox{as } x \to \infty. \, Those that are two-tailed include: 有两条尾巴的包括: }[/math]


This has the intuitive interpretation for a right-tailed long-tailed distributed quantity that if the long-tailed quantity exceeds some high level, the probability approaches 1 that it will exceed any other higher level.


All long-tailed distributions are heavy-tailed, but the converse is false, and it is possible to construct heavy-tailed distributions that are not long-tailed.


Subexponential distributions

A fat-tailed distribution is a distribution for which the probability density function, for large x, goes to zero as a power x^{-a}. Since such a power is always bounded below by the probability density function of an exponential distribution, fat-tailed distributions are always heavy-tailed. Some distributions, however, have a tail which goes to zero slower than an exponential function (meaning they are heavy-tailed), but faster than a power (meaning they are not fat-tailed). An example is the log-normal distribution . Many other heavy-tailed distributions such as the log-logistic and Pareto distribution are, however, also fat-tailed.

肥尾分布是一个分布,对于大的 x,概率密度函数变为0的 x ^ {-a }的幂。由于这样的幂总是以概率密度函数的指数分布为界,胖尾分布总是重尾分布。然而,有些分布有一条尾巴,它比指数函数分布慢到零(意味着它们是重尾分布) ,但比幂分布快(意味着它们不是厚尾分布)。对数正态分布就是一个例子。然而,许多其他的重尾分布,例如 log-logistic 分布和帕累托分布分布也是厚尾分布。


Subexponentiality is defined in terms of convolutions of probability distributions. For two independent, identically distributed random variables [math]\displaystyle{ X_1,X_2 }[/math] with common distribution function [math]\displaystyle{ F }[/math] the convolution of [math]\displaystyle{ F }[/math] with itself, [math]\displaystyle{ F^{*2} }[/math] is convolution square, using Lebesgue–Stieltjes integration, by:


[math]\displaystyle{ There are parametric (see Embrechts et al.) approaches to the problem of the tail-index estimation. 有参数(参见 Embrechts 等人。)对尾部指数估计问题的探讨。 \Pr[X_1+X_2 \leq x] = F^{*2}(x) = \int_{0}^x F(x-y)\,dF(y), }[/math]

To estimate the tail-index using the parametric approach, some authors employ GEV distribution or Pareto distribution; they may apply the maximum-likelihood estimator (MLE).

为了使用参数方法估计尾部指数,一些作者使用了 GEV 分布或帕累托分布,他们可以使用最大似然估计(MLE)。

and the n-fold convolution [math]\displaystyle{ F^{*n} }[/math] is defined inductively by the rule:

[math]\displaystyle{ F^{*n}(x) = \int_{0}^x F(x-y)\,dF^{*n-1}(y). }[/math]

With (X_n , n \geq 1) a random sequence of independent and same density function F \in D(H(\xi)), the Maximum Attraction Domain of the generalized extreme value density H , where \xi \in \mathbb{R}. If \lim_{n\to\infty} k(n) = \infty and \lim_{n\to\infty} \frac{k(n)}{n}= 0, then the Pickands tail-index estimation is

在 d (h (xi))中具有独立相等密度函数 f 的随机序列,得到了广义极值密度 h 的最大吸引域,其中 xi 在 mathbb { r }中。如果 lim _ { n to infty } k (n) = infty,lim _ { n to infty } frac { k (n)}{ n } = 0,则 Pickands 尾指数估计为


The tail distribution function [math]\displaystyle{ \overline{F} }[/math] is defined as [math]\displaystyle{ \overline{F}(x) = 1-F(x) }[/math].

[math]\displaystyle{ 

《数学》



\xi^\text{Hill}_{(k(n),n)} = \left(\frac 1 {k(n)} \sum_{i=n-k(n)+1}^n \ln(X_{(i,n)}) - \ln (X_{(n-k(n)+1,n)})\right)^{-1},

1{ n)} = left (frac 1{ k (n)} sum { i = n-k (n) + 1} ^ n ln (x _ { i,n)})-ln (x _ { n-k (n) + 1,n)}) right) ^ {-1} ,

A distribution \lt math\gt F }[/math] on the positive half-line is subexponential[1][4][5] if

</math>

数学


[math]\displaystyle{ where X_{(i,n)} is the i-th order statistic of X_1, \dots, X_n. 其中 x _ {(i,n)}是 x _ 1,点,x _ n 的 i 阶统计量。 \overline{F^{*2}}(x) \sim 2\overline{F}(x) \quad \mbox{as } x \to \infty. This estimator converges in probability to \xi, and is asymptotically normal provided k(n) \to \infty is restricted based on a higher order regular variation property 该估计量在概率上收敛到 xi,并且基于高阶正则变差性质,在 k (n)为信度的条件下,它是渐近正态的 }[/math]

. Consistency and asymptotic normality extend to a large class of dependent and heterogeneous sequences, irrespective of whether X_t is observed, or a computed residual or filtered data from a large class of models and estimators, including mis-specified models and models with errors that are dependent.

.相合性和渐近正态性扩展到一大类相依和异质序列,无论是否观测到 x _ t,还是来自一大类模型和估计器的计算残差或过滤数据,包括误差相依的模型和模型。


This implies[6] that, for any [math]\displaystyle{ n \geq 1 }[/math],


[math]\displaystyle{ The ratio estimator (RE-estimator) of the tail-index was introduced by Goldie 引入了尾部指数的比率估计量(re- 估计量) \overline{F^{*n}}(x) \sim n\overline{F}(x) \quad \mbox{as } x \to \infty. and Smith. 还有史密斯。 }[/math]

It is constructed similarly to Hill's estimator but uses a non-random "tuning parameter".

它的构造类似于希尔估计器,但使用了一个非随机的“调谐参数”。


The probabilistic interpretation[6] of this is that, for a sum of [math]\displaystyle{ n }[/math] independent random variables [math]\displaystyle{ X_1,\ldots,X_n }[/math] with common distribution [math]\displaystyle{ F }[/math],

A comparison of Hill-type and RE-type estimators can be found in Novak.

希尔型和稀土型估计量的比较可以在 Novak 找到。


[math]\displaystyle{ \Pr[X_1+ \cdots +X_n\gt x] \sim \Pr[\max(X_1, \ldots,X_n)\gt x] \quad \text{as } x \to \infty. }[/math]

Nonparametric approaches to estimate heavy- and superheavy-tailed probability density functions were given in

给出了估计重尾和超重尾概率密度函数的非参数方法


Markovich. These are approaches based on variable bandwidth and long-tailed kernel estimators; on the preliminary data transform to a new random variable at finite or infinite intervals which is more convenient for the estimation and then inverse transform of the obtained density estimate; and "piecing-together approach" which provides a certain parametric model for the tail of the density and a non-parametric model to approximate the mode of the density. Nonparametric estimators require an appropriate selection of tuning (smoothing) parameters like a bandwidth of kernel estimators and the bin width of the histogram. The well known data-driven methods of such selection are a cross-validation and its modifications, methods based on the minimization of the mean squared error (MSE) and its asymptotic and their upper bounds. A discrepancy method which uses well-known nonparametric statistics like Kolmogorov-Smirnov's, von Mises and Anderson-Darling's ones as a metric in the space of distribution functions (dfs) and quantiles of the later statistics as a known uncertainty or a discrepancy value can be found in.

男名男子名。这些方法包括: 基于变带宽和长尾核估计的方法; 在有限或无限区间内将初始数据转换为一个新的随机变量的方法,这种方法更便于对所得密度估计进行估计和反变换; 以及“拼接方法” ,这种方法为密度的尾部提供了一个确定的参数模型和一个非参数模型来逼近密度的模式。非参数估计需要适当选择调整(平滑)参数,如核估计的带宽和直方图的容器宽度。众所周知的数据驱动选择方法是交叉验证及其修正,基于最小均方差及其渐近和上界的方法。在分布函数空间(dfs)和后续统计量的分位数空间(dfs)中,利用著名的无母数统计,如 Kolmogorov-Smirnov、 von Mises 和 Anderson-Darling 的分布函数,作为已知的不确定性或不一致值,可以找到一种差异方法。

This is often known as the principle of the single big jump[7] or catastrophe principle.[8]


A distribution [math]\displaystyle{ F }[/math] on the whole real line is subexponential if the distribution

[math]\displaystyle{ F I([0,\infty)) }[/math] is.[9] Here [math]\displaystyle{ I([0,\infty)) }[/math] is the indicator function of the positive half-line. Alternatively, a random variable [math]\displaystyle{ X }[/math] supported on the real line is subexponential if and only if [math]\displaystyle{ X^+ = \max(0,X) }[/math] is subexponential.


All subexponential distributions are long-tailed, but examples can be constructed of long-tailed distributions that are not subexponential.


Common heavy-tailed distributions

All commonly used heavy-tailed distributions are subexponential.[6]


Those that are one-tailed include:

Category:Tails of probability distributions

类别: 概率分布的尾部

Category:Types of probability distributions

类别: 概率分布的类型

Category:Actuarial science

类别: 精算

Category:Risk

类别: 风险


This page was moved from wikipedia:en:Heavy-tailed distribution. Its edit history can be viewed at 重尾分布/edithistory

  1. 1.0 1.1 1.2 Asmussen, S. R. (2003). "Steady-State Properties of GI/G/1". Applied Probability and Queues. Stochastic Modelling and Applied Probability. 51. pp. 266–301. doi:10.1007/0-387-21525-5_10. ISBN 978-0-387-00211-8. 
  2. 2.0 2.1 Rolski, Schmidli, Scmidt, Teugels, Stochastic Processes for Insurance and Finance, 1999
  3. S. Foss, D. Korshunov, S. Zachary, An Introduction to Heavy-Tailed and Subexponential Distributions, Springer Science & Business Media, 21 May 2013
  4. Chistyakov, V. P. (1964). "A Theorem on Sums of Independent Positive Random Variables and Its Applications to Branching Random Processes". ResearchGate (in English). Retrieved April 7, 2019.
  5. Teugels, Jozef L. (1975). "The Class of Subexponential Distributions". University of Louvain: Annals of Probability. Retrieved April 7, 2019.
  6. 6.0 6.1 6.2 Embrechts P.; Klueppelberg C.; Mikosch T. (1997). Modelling extremal events for insurance and finance. Stochastic Modelling and Applied Probability. 33. Berlin: Springer. doi:10.1007/978-3-642-33483-2. ISBN 978-3-642-08242-9. 
  7. Foss, S.; Konstantopoulos, T.; Zachary, S. (2007). "Discrete and Continuous Time Modulated Random Walks with Heavy-Tailed Increments" (PDF). Journal of Theoretical Probability. 20 (3): 581. arXiv:math/0509605. CiteSeerX 10.1.1.210.1699. doi:10.1007/s10959-007-0081-2.
  8. Wierman, Adam (January 9, 2014). "Catastrophes, Conspiracies, and Subexponential Distributions (Part III)". Rigor + Relevance blog. RSRG, Caltech. Retrieved January 9, 2014.
  9. Willekens, E. (1986). "Subexponentiality on the real line". Technical Report. K.U. Leuven.