摘要:

We derive an estimator for the time of change in a linear model when the regression coefficients and the variance may change. The asymptotic distribution of the estimator is computed. The method is used to construct confidence interval for the time of change in gross domestic product in the United States between 1929 and 1967.

ISSN:0233-1888    

DOI:10.1080/02331889708802578

出版商:Gordon & Breach Science Publishers    

年代:1997

数据来源: Taylor

ISSN:0233-1888    

DOI:10.1080/02331889708802579

出版商:Gordon & Breach Science Publishers    

年代:1997

数据来源: Taylor

摘要:

The trace of the scale matrix of the multivariatet-distribution is considered for estimation using a regression type estimator. The proposed estimator always dominates the usual maximum likelihood estimator under a quadratic loss function. The dominance behaviour is proved analytically as well as empirically using Monte-Carlo simulation.

ISSN:0233-1888    

DOI:10.1080/02331889708802580

出版商:Gordon & Breach Science Publishers    

年代:1997

数据来源: Taylor

摘要:

Gajek and Gather(1991) showed that under the presence of censoring a trivial estimator of the scale parameter not depending on the sample can be both minimax and admissible. We consider a general decision problem. No assumptions are imposed on decision space, decision procedures and loss function except standard ones. It is shown that the anomaly also occurs. Some related results are given.

ISSN:0233-1888    

DOI:10.1080/02331889708802581

出版商:Gordon & Breach Science Publishers    

年代:1997

数据来源: Taylor

摘要:

This paper presents some models of markings and thinnings of Poisson processes. Examples of marks are distances and directions to nearest neighbours. The concept of the mark variogram is introduced. Using a simple relation between thinned Poisson processes, the second-order product density of that cluster process is given which is complementary to Matern’s hard core process. An example shows that it is of value for spatial statistics.

ISSN:0233-1888    

DOI:10.1080/02331889708802582

出版商:Gordon & Breach Science Publishers    

年代:1997

数据来源: Taylor

ISSN:0233-1888    

DOI:10.1080/02331889708802583

出版商:Gordon & Breach Science Publishers    

年代:1997

数据来源: Taylor

ISSN:0233-1888    

DOI:10.1080/02331889708802577

出版商:Gordon & Breach Sceince Publishers    

年代:1997

数据来源: Taylor