Titlebook: Breakthroughs in Statistics; Samuel Kotz,Norman L. Johnson Book 1997 Springer Science+Business Media New York 1997 algorithms.boundary ele

憤怒事實

nonradioactive

, Daniels (1954) Saddlepoint Approximations in Statistics,nt techniques in statistics. These have proved to be a general and valuable tool for deriving very accurate approximations to the distribution of general estimators and test statistics both in parametric and nonparametric situations. Secondly, it shows the relationship between saddlepoint approximat

放牧

,, Hájek (1970) A Characterization of Limiting Distributions of Regular Estimates,tion theorem emerged as structure hidden behind the asymptotic informa-tion bound. In recent work, the convolution theorem has re-emerged as structure underlying correct convergence of bootstrap distributions. Both bootstrapping and the information inequality deceive naive users at the superefficien

代理人

Nomadic

不能強(qiáng)迫我

, Besag (1974) Spatial Interaction and the Statistical Analysis of Lattice Systems,ract serious attention. Thus Besag (1986) was the second major paper [after Geman and Geman (1984), featured elsewhere in this volume] on the Markov random fields approach to image analysis, while in the 1990s he made a number of contributions to Markov chain Monte Carlo (MCMC) sampling [e.g., Besag

puzzle

, Levit (1974) On the Optimality of Some Statistical Estimates,ity as a mathematical theorem, at least asymptotically. A first result in this direction is due to Edgeworth (1908), who proves the maximum likelihood estimator of a location parameter to be asymptotically optimal among all estimators which are solutions to estimating equations. Various attempts by