作者: Jaroslav H�jek
DOI: 10.1007/BF00533669
关键词:
摘要: We consider a sequence of estimates in general estimation problems with k-dimensional parameter. Under certain very conditions we prove that the limiting distribution estimates, if properly normed, is convolution normal distribution, which depends only underlying distributions, and further on choice estimate. As corollaries obtain inequalities for asymptotic variances probabilities sets, generalizing so some results J. Wolfowitz (1965), S. Kaufman (1966), L. Schmetterer (1966) G. Roussas (1968).