M-估计的强相合性

The strong consistence of M-estimate

本文在x1,x2,…,xn,…为非随机变量的情况下讨论了模型yi=f(xi,θ)+εi i=1,2,…,n下的M—估计■n的强相合性,其中■n满足nΣi=1p(yi-f(xi,■n))=minθ∈⊙nΣi=1p(yi-f(xi,θ))给出了M-估计具有强相合性的一个充分条件.

In this paper ,we study the strong consistence of M-estimate from the minimizing problem ∑^ni=1p（yi-f（xi,θn））=minθ∈⊙ ∑^ni=1p（yi-f（xi,θ）） where X1 X2,… .are the observations of a q-dimension unrandom vecters x. We get a sufllcient condition about the strong consistence of M-estimate.