「標準」分布Fによって決定されるロケーションスケールファミリを検討します。F
ΩF={F(μ,σ):x→F(x−μσ)∣σ>0}.
微分可能と仮定すると、PDFは1F。1σf((x−μ)/σ)dx
これらの配布を切り捨て、aとbの間のサポートを制限するため、a < bは、PDFが次のように置き換えられることを意味します。aba<b
f(μ,σ;a,b)(x)=f(x−μσ)dxσC(μ,σ,a,b),a≤x≤b
(他のすべての値ではゼロです)ここで、C (μ 、σ 、a 、b )= F (μ 、σ )(b )− F (μ 、σ )(a )は、f (μ 、σ ; a 、b )は単一性に統合されます。(Cはまったく同じであることに注意してください1xC(μ,σ,a,b)=F(μ,σ)(b)−F(μ,σ)(a)f(μ,σ;a,b)C1 in the absence of truncation.) The log likelihood for iid data xi therefore is
Λ(μ,σ)=∑i[logf(xi−μσ)−logσ−logC(μ,σ,a,b)].
Critical points (including any global minima) are found where either σ=0 (a special case I will ignore here) or the gradient vanishes. Using subscripts to denote derivatives, we may formally compute the gradient and write the likelihood equations as
00=∂Λ∂μ=∂Λ∂σ=∑i⎡⎣⎢−fμ(xi−μσ)f(xi−μσ)−Cμ(μ,σ,a,b)C(μ,σ,a,b)⎤⎦⎥=∑i⎡⎣⎢−fσ(xi−μσ)σ2f(xi−μσ)−1σ−Cσ(μ,σ,a,b)C(μ,σ,a,b)⎤⎦⎥
Because a and b are fixed, drop them from the notation and write nCμ(μ,σ,a,b)/C(μ,σ,a,b) as A(μ,σ) and nCσ(μ,σ,a,b)/C(μ,σ,a,b) as B(μ,σ). (With no truncation, both functions would be identically zero.) Separating the terms involving the data from the rest gives
−A(μ,σ)−σ2B(μ,σ)−nσ=∑ifμ(xi−μσ)f(xi−μσ)=∑ifσ(xi−μσ)f(xi−μσ)
By comparing these to the no-truncation situation it is evident that
Any sufficient statistics for the original problem are sufficient for the truncated problem (because the right hand sides have not changed).
Our ability to find closed-form solutions relies on the tractability of A and B. If these do not involve μ and σ in simple ways, we cannot hope to obtain closed-form solutions in general.
For the case of a normal family, C(μ,σ,a,b) of course is given by the cumulative normal PDF, which is a difference of error functions: there is no chance that a closed-form solution can be obtained in general. However, there are only two sufficient statistics (the sample mean and variance will do) and the CDF is as smooth as can be, so numerical solutions will be relatively easy to obtain.