ISSN:0040-1706    

DOI:10.1080/00401706.1989.10488469

出版商:Taylor & Francis Group    

年代:1989

数据来源: Taylor

摘要:

A simple method is presented for fitting regression models that are nonlinear in the explanatory variables. Despite its simplicity—or perhaps because of it—the method has some powerful characteristics that cause it to be competitive with and often superior to more sophisticated techniques, especially for small data sets in the presence of high noise.

ISSN:0040-1706    

DOI:10.1080/00401706.1989.10488470

出版商:Taylor & Francis Group    

年代:1989

数据来源: Taylor

ISSN:0040-1706    

DOI:10.1080/00401706.1989.10488471

出版商:Taylor & Francis Group    

年代:1989

数据来源: Taylor

ISSN:0040-1706    

DOI:10.1080/00401706.1989.10488472

出版商:Taylor & Francis Group    

年代:1989

数据来源: Taylor

ISSN:0040-1706    

DOI:10.1080/00401706.1989.10488473

出版商:Taylor & Francis Group    

年代:1989

数据来源: Taylor

摘要:

A computer experiment generates observations by running a computer model at inputsxand recording the output (response)Y. Prediction of the responseYto an untried input is treated by modeling the systematic departure ofYfrom a linear model as a realization of a stochastic process. For given data (selected inputs and the computed responses), best linear prediction is used. The design problem is to select the inputs to predict efficiently. The issues of choice of stochastic-process model and computation of efficient designs are addressed, and applications are made to some chemical kinetics problems.

ISSN:0040-1706    

DOI:10.1080/00401706.1989.10488474

出版商:Taylor & Francis Group    

年代:1989

数据来源: Taylor

摘要:

This article summarizes recent work in optimal experimental design in nonlinear problems, in which the major difficulty in obtaining good or optimal designs is their dependence on the true value of the parameters. This difficulty arises in problems with nonlinear models or with linear models in which interest lies in a nonlinear function of the parameters. Most approaches use a static design based on “prior” information about the parameters or a sequential procedure that takes advantage of the inflow of new information about them. The various versions of these methods are discussed, as are some of the consequent problems of inference. Some selected procedures are compared using simulation studies.

ISSN:0040-1706    

DOI:10.1080/00401706.1989.10488475

出版商:Taylor & Francis Group    

年代:1989

数据来源: Taylor

摘要:

The sorted binary plot is a graphical method for identifying and displaying patterns in multivariate data sets. The construction requires calculation of the median for each variable measured, followed by subtraction of the medians from the values for each sample. The signs of the residuals represent a binary number for each sample. The list of binary numbers is sorted and converted to a graph by assigning distinctive symbols to 1 and 0. The sorting operation causes samples with the same binary number to form clusters. The method can be extended to three or more quantiles. It has been applied to several kinds of data.

ISSN:0040-1706    

DOI:10.1080/00401706.1989.10488476

出版商:Taylor & Francis Group    

年代:1989

数据来源: Taylor

摘要:

This article describes the use of bootstrap methods for the problem of testing homogeneity of variances when means are not assumed equal or known. The methods are new in this context and allow the use of normal-theory test statistics such asF=s21/s22without the normality assumption that is crucial for validity of critical values obtained from theFdistribution. Both asymptotic analysis and Monte Carlo sampling show that the new resampling procedures compare favorably with older methods in terms of test validity and power.

ISSN:0040-1706    

DOI:10.1080/00401706.1989.10488477

出版商:Taylor & Francis Group    

年代:1989

数据来源: Taylor

摘要:

Changes in calibration curves from one time to the next, caused by drift, often require measuring devices to be recalibrated at frequent intervals. In such situations the usual practice is to estimate the unknown values of test samples using only data from the corresponding calibration period. Under a random coefficient regression model for the different calibration curves, however, it can be shown that it is more efficient to combine the data from all calibration periods to estimate the unknowns. We consider a particular class of point estimators obtained by inverting suitable prediction functions and show that the estimator obtained from a best prediction function is optimal in a sense defined by Godambe (1960) and Durbin (1960) in the context of unbiased estimating equations. We also compare the smallsample performance of this estimator with the usual estimator using the Pitman closeness criterion.

ISSN:0040-1706    

DOI:10.1080/00401706.1989.10488478

出版商:Taylor & Francis Group    

年代:1989

数据来源: Taylor