指数加重移動歪度/尖度


15

指数加重移動平均と標準プロセスの偏差を計算するためのよく知られたオンライン式がある(xn)n=0,1,2,…。平均して、

μn=(1−α)μn−1+αxn

分散について

σn2=(1−α)σn−12+α(xn−μn−1)(xn−μn)

ここから標準偏差を計算できます。

指数加重された第3および第4中心モーメントのオンライン計算のための同様の公式はありますか?私の直感は、彼らが形をとるべきだということです

M3,n=(1−α)M3,n−1+αf(xn,μn,μn−1,Sn,Sn−1)

そして

M4,n=(1−α)M4,n−1+αf(xn,μn,μn−1,Sn,Sn−1,M3,n,M3,n−1)

あなたは歪度計算することができたからおよび尖度K N = M 4 、N / σ 4 Nが、私は、機能のための閉形式の単純なを見つけることができるされていませんでしたFとG。γn=M3,n/σn3kn=M4,n/σn4fg


編集:いくつかの詳細情報。移動分散の更新式は、指数加重移動共分散の式の特殊なケースです。

Cn(x,y)=(1−α)Cn−1(x,y)+α(xn−x¯n)(yn−y¯n−1)

こことˉ Y Nの指数関数的な移動手段であるX及びYは。間の非対称のxとyの錯覚であり、あなたがそれに気づくとき消えYが- ˉ Y N = (1 - α )(Y - ˉ Y N - 1)。x¯ny¯nxyxyy−y¯n=(1−α)(y−y¯n−1)

このような式は、中心モーメントを期待値として記述し、期待値の重みが指数関数的であると理解し、関数f (x )に対してEn(⋅)f(x)

En(f(x))=αf(xn)+(1−α)En−1(f(x))

It's easy to derive the updating formulas for the mean and variance using this relation, but it's proving to be more tricky for the third and fourth central moments.

回答:


6

The formulas are straightforward but they are not as simple as intimated in the question.

Let Y be the previous EWMA and let X=xn, which is presumed independent of Y. By definition, the new weighted average is Z=αX+(1−α)Y for a constant value α. For notational convenience, set β=1−α. Let F denote the CDF of a random variable and ϕ denote its moment generating function, so that

ϕX(t)=EF[exp⁡(tX)]=∫Rexp⁡(tx)dFX(x).

With Kendall and Stuart, let μk′(Z) denote the non-central moment of order k for the random variable Z; that is, μk′(Z)=E[Zk]. The skewness and kurtosis are expressible in terms of the μk′ for k=1,2,3,4; for example, the skewness is defined as μ3/μ23/2 where

μ3=μ3′−3μ2′μ1′+2μ1′3 and μ2=μ2′−μ1′2

are the third and second central moments, respectively.

By standard elementary results,

1+μ1′(Z)t+12!μ2′(Z)t2+13!μ3′(Z)t3+14!μ4′(Z)t4+O(t5)=ϕZ(t)=ϕαX(t)ϕβY(t)=ϕX(αt)ϕY(βt)=(1+μ1′(X)αt+12!μ2′(X)α2t2+⋯)(1+μ1′(Y)βt+12!μ2′(Y)β2t2+⋯).

To obtain the desired non-central moments, multiply the latter power series through fourth order in t and equate the result term-by-term with the terms in ϕZ(t).


I am having some formula visualization problem, possibly whenever a ' is used, with both IE and Firefox, would you please care checking? Thanks!
— Quartz

1
@Quartz Thanks for the heads up. This used to display properly, so evidently there has been some change in the processing of the TEX markup. I found a workaround by enclosing all single quotes within braces. (This change has probably broken a few dozen posts on this site.)
— whuber

0

I think that the following updating formula works for the third moment, although I'd be glad to have someone check it:

M3,n=(1−α)M3,n−1+α[xn(xn−μn)(xn−2μn)−xnμn−1(μn−1−2μn)−… ⋯−μn−1(μn−μn−1)2−3(xn−μn)σn−12]

Updating formula for the kurtosis still open...


Why the ... in the above formula?
— Chris

Line continuation.
— Chris Taylor

Did your equation prove to be correct? I asked a similar question in R. stats.stackexchange.com/q/234460/70282
— Chris

Did you account for the division by N in the third moment? Skewness is the ratio of the 3rd moment and the standard deviation^3 like so: Skew = m3 / sqrt(variance)^3 The third moment is defined as: m3 = sum( (x-mean)^3 )/n
— Chris
弊社のサイトを使用することにより、あなたは弊社のクッキーポリシーおよびプライバシーポリシーを読み、理解したものとみなされます。
Licensed under cc by-sa 3.0 with attribution required.