摘要:

This paper deals with the sequential point estimation of the autoregressive parameters in a multiple autoregressive model using the least squares estimator.The sequential estimator is shown to be asymptotically risk efficient under some regularity conditions.The asymptotic normality and uniform integrability of standardized stopping rule are established.This paper also contain a second order approximation to expected stopping time and an expression for regret of the above stopping rule under certain smoothness conditions.

ISSN:0747-4946    

DOI:10.1080/07474949708836370

出版商:Marcel Dekker, Inc.    

年代:1997

数据来源: Taylor

摘要:

The paper considers a sequential estimator for the parameter of the first order autoregressive model with unknown noise variance based on the generalized least squares estimator. The estimator has a prescribed mean squared accuracy and is asymptotically normally distributed uniformly in the autoregressive parameter taking values in any compact subset of the real line. The estimator is shown to be asymptotically minimax (in the Hajek-LeCam sense) in a large class of sequential and fixed sample size esti~nation procedures.

ISSN:0747-4946    

DOI:10.1080/07474949708836371

出版商:Marcel Dekker, Inc.    

年代:1997

数据来源: Taylor

摘要:

A sequential estimation problem on a linearly combined parameter η := aθ+ bξ for a family of distributions with two parameters θ and ξ is considered. The Bhattacharyya type bound for the asymptotic variance of sequential estimators of η in the presence of a linearly combined nuisance parameter is given up to the second order and its attainment is also discussed. Further it is shown that the modified maximum likelihood estimation procedure of η with an appropriate stopping rule attains the bound for a family of normal distributions.

ISSN:0747-4946    

DOI:10.1080/07474949708836372

出版商:Marcel Dekker, Inc.    

年代:1997

数据来源: Taylor

摘要:

In the context of comparing two treatments in clinical trials randomized play-the-winner rule is used with the goal of treating more patients by the better treatment On its basis we give two decision rules by introducing stopping rule Some performance characteristics of one of the decision rules are discussed and examined .A minimax rule is suggested Finally some exact and asymptotic properties of the decision rule is derived

ISSN:0747-4946    

DOI:10.1080/07474949708836373

出版商:Marcel Dekker, Inc.    

年代:1997

数据来源: Taylor

摘要:

An economical on-line quality control procedure is given for cases where measurement error u present.The unobservable quality characteristic procw is assumed to follow a normal random walk.model independent of the measurement error which is white noise. The off-target loss function used is proportional to the squared deviation from the target value.The procw is inspected at regular intervals of m time units and adjusted at the end of a cycle when the process goes out of the control limit d.The limning long-run average cost rate is evaluated.The optimum value of the control parameters,the inspection interval m,and the control limit d are obtained by minimlzlng the long-run average cost rate.When m is small and d is large, explicit expressions for the control parameters are given.

ISSN:0747-4946    

DOI:10.1080/07474943708836374

出版商:Marcel Dekker, Inc.    

年代:1997

数据来源: Taylor

摘要:

This paper considers sequantial producers to construct fixed-width confidence intervals for some function θ of mean μ and variance σ2of normal distribution.Consideration is devoted to θ=exp( μ + σ2/2 ) and θ=μ/σ.Nonlinear renewal theory is used to drive asymptotic expansion of expectation of the stopping time and the estimate as the width of confidence interval decreases to zero.An improvement of the coverage probability is also discussed.

ISSN:0747-4946    

DOI:10.1080/07474949708836375

出版商:Marcel Dekker, Inc.    

年代:1997

数据来源: Taylor

摘要:

Suppose we have a function h with m arguments and i.i.d. random variableswith marginal distribution F. Let Hrbe the distribution of einf:,...,Xm), m≥ 2. We consider on-line schemes for estimating quantiles of HF. Such an estimator is based on a design Dn, which is a small subset of all n!/(n-m)! possible index vectors I = (il) having distinct entries not exceeding n. When a new observation Xnarrivesnew vectorswithare used to modify the current estimate. When γ → ∞, the asymptotic relative efficiency of the recursive estimator compared to the off-line estimator (U -quantile) tends to one. The on-line estimator is closely related to incomplete U-quantiles (Hössjer, 1996), and it generalizes a recursive quantile estimator considered by Hoist (1987) for m = 1.

ISSN:0747-4946    

DOI:10.1080/07474949708836376

出版商:Marcel Dekker, Inc.    

年代:1997

数据来源: Taylor

ISSN:0747-4946    

DOI:10.1080/07474949708836369

出版商:Marcel Dekker, Inc.    

年代:1997

数据来源: Taylor