ISSN:0162-1459    

DOI:10.1080/01621459.1997.10473597

出版商:Taylor & Francis Group    

年代:1997

数据来源: Taylor

摘要:

Multiple outcomes are commonly measured in the study of birth defects. The reason is that most teratogens do not cause a single, uniquely defined defect, but rather result in a range of effects, including major malformations, minor anomalies, and deficiencies in birth weight, length and head circumference. The spectrum of effects associated with a particular teratogen is sometimes described as a “syndrome.” In this article we develop a latent variable model to characterize exposure effects on multiple binary outcomes. Not only does the method allow comparisons of control and exposed infants with respect to multiple outcomes, but it also provides a measure of the “severity” of each child's condition. Data from a study of the teratogenic effects of anticonvulsants illustrate our results.

ISSN:0162-1459    

DOI:10.1080/01621459.1997.10473598

出版商:Taylor & Francis Group    

年代:1997

数据来源: Taylor

摘要:

Epidemiologists commonly study the geographical variation of disease rates to generate and refine testable hypotheses regarding etiology. Poisson regression is often used to estimate the parameters that characterize the effects of risk factors on disease. Hierarchical models have been proposed to utilize spatial locations and neighbors as surrogates for unknown or unmeasured risk factors in the analysis of disease rates. Although hierarchical models are useful in modeling spatial disease rates in a scientifically meaningful way, the analytic tools for them are generally computationally intensive. We overcome this issue by applying a conditional spatial modeling technique using unbiased estimating functions. The resulting estimator of regression coefficients is consistent and asymptotically normally distributed under mild conditions. A simulation study compares the performance of the proposed scheme with the approximate inference methods for hierarchical models. We apply the proposed scheme to Scottish lip cancer incidence data to illustrate its use in practice.

ISSN:0162-1459    

DOI:10.1080/01621459.1997.10473599

出版商:Taylor & Francis Group    

年代:1997

数据来源: Taylor

摘要:

The management of a commercial fishery often requires estimates of the age composition. These estimates are typically based on age and length data obtained from sampling the commercial landings from the fishery and the catches from a research vessel survey. We use data from annual research cruises of Georges Bank conducted by the Canadian Department of Fisheries and Oceans to show how inferences for Atlantic cod (Gadus morhua) can be improved. Traditionally, two-phase stratified sampling is used with fish length (or weight) as the stratification variable. LettingPi. denote the proportion of fish belonging to length stratumi, and πijdenote the proportion of fish belonging to age classjin stratumi, we use Bayesian methods to estimateP.j= πiπij, the proportion of fish that are agej.Specifications of smoothness, expressed as unimodal order relations among the πij(within and between the length strata), are incorporated into the prior distributions. Uncertainty about both the locations of the modes and the unimodality itself are included as part of the probabilistic specification. With computations facilitated by using the Gibbs sampler, we show that the smoothness conditions provide very large gains in precision. For the data analyzed in this article, one can obtain similar precision by using (a) a conventional analysis or (b) an analysis with order restrictions and a sample of half the size in (a). We also show that better estimates of the age composition provide improved estimates of the quantities used by fisheries managers to forecast the catches from cohorts of fish.

ISSN:0162-1459    

DOI:10.1080/01621459.1997.10473600

出版商:Taylor & Francis Group    

年代:1997

数据来源: Taylor

摘要:

The authorship of portions of some ancient and medieval astronomical manuscripts is a current historical issue. This article develops a statistical method for determining whether or not a link exists between a pair of mathematically defined historical astronomical tables. The existence or nonexistence of such links aids the analysis of the origins of these documents. Nonparametric methods are used to determine whether traces of the errors in one table may be found in the other, or whether the “dependent” table is too accurate to have been derived from the entries in the other table. Several applications reveal new historical information about certain tables in Claudius Ptolemy'sAlmagest(c. A.D. 140) and about the tables of the 14th-century Islamic astronomer Shams al-Dɨn al-Khalɨlɨ.

ISSN:0162-1459    

DOI:10.1080/01621459.1997.10473601

出版商:Taylor & Francis Group    

年代:1997

数据来源: Taylor

摘要:

With suitably chosen priors, Bayesian models are useful in image reconstruction. I consider reconstructions from single-photon emission computerized tomography data using a Gibbs pairwise difference prior. A fully Bayesian approach is presented where the prior parameters are considered drawn from hyperpriors. The approach is problematical, because the normalization constant in the prior is an intractable function of its parameters. Here the constant is estimated off-line by reverse logistic regression. Markov chain Monte Carlo methods are used on simulated and real data to gain estimates of the posterior mean and pixelwise credibility bounds. The resulting reconstructions are compared to those obtained by filtered back-projection, maximum likelihood/EM, and the one-step-late solution to the fixed parameter posterior mode.

ISSN:0162-1459    

DOI:10.1080/01621459.1997.10473602

出版商:Taylor & Francis Group    

年代:1997

数据来源: Taylor

摘要:

We address the commonly encountered situation in spatial statistics where data such as counts of incidences of a certain disease are available only in geographically aggregated form. We develop fairly general models and propose a modified version of the locally weighted least squares method to recover the unknown smooth spatial function that is assumed to generate the observations. In the special case of count data, the target function is the intensity function, conditional on the total number of observations. Our method avoids the arbitrariness of selecting a point within each geographic area at which the measurement for the whole area is supposed to be located. We derive basic asymptotic properties, and apply our methods to acquired immune deficiency syndrome (AIDS) incidence data in San Francisco for 1980–1992, where counts are available aggregated over zip code areas.

ISSN:0162-1459    

DOI:10.1080/01621459.1997.10473603

出版商:Taylor & Francis Group    

年代:1997

数据来源: Taylor

摘要:

We propose a nonparametric method of data analysis for situations where the data consist of a sample of curves and a covariate is present. We assume the observed curves to be time-accelerated versions of a basic underlying stochastic process and assume the time acceleration factor for each observed process, determining the “eigenzeit,” to be a smooth function of the covariate. We discuss an iterative procedure to estimate the components of the model nonparametrically, using cross-validation to determine smoothing parameters using a leave-one-curve-out technique. Our example concerns time courses of mortality for a sample of cohorts of fruit flies.

ISSN:0162-1459    

DOI:10.1080/01621459.1997.10473604

出版商:Taylor & Francis Group    

年代:1997

数据来源: Taylor

摘要:

In problems of medical imaging and robotic vision, measurements of a convex set are sometimes obtained via the set's support function. The standard way in which the convex set is recovered from such data is to suppose that it is polygonal, model the errors as Normal random variables, and apply constrained maximum likelihood methods. Typically, the number of sides assumed of the polygon is equal to or a little less than the number of data points. However, from a statistical viewpoint, the number of sides should really be interpreted as a smoothing parameter and chosen to optimize some measure of performance. Additionally, if the true set is not a polygon, then a polygonal estimate can be aesthetically unsatisfactory. In this article we suggest periodic smoothing methods for estimating the convex set.

ISSN:0162-1459    

DOI:10.1080/01621459.1997.10473605

出版商:Taylor & Francis Group    

年代:1997

数据来源: Taylor

摘要:

Segmentation has become a widely used tool in modern image analysis. Typical methods are highly nonlinear and often ad hoc, and none so far have proved accessible to detailed statistical analysis even by asymptotic approximations. Can such methods be any good? This article describes a segmentation procedure for multichannel image data and attempts to develop an understanding of its statistical performance characteristics in terms of nonparametric regression. The procedure is based on a recursive merging algorithm defined via a nested sequence of discretizations of the image domain. A cross-validation rule with a near-neighbor block replacement strategy is proposed for selecting the final segmentation model. Idealized numerical simulation experiments are used to evaluate the rate of convergence of the mean square error of estimation and also to study the efficiency of the cross-validation technique. Interestingly, the results show that the rate of convergence tends to decrease as the degree of smoothness of the underlying image increases. This complements the more familiar estimation characteristic associated with conventional nonparametric regression algorithms. The cross-validation is found to be effective at choosing a segmentation that minimizes the mean squared deviation between the segmented image and the underlying truth. Furthermore, by manipulating the block size in the replacement scheme, it is possible to maintain some robustness to artifacts caused by blurring. Physical phantom datasets taken from positron emission tomography (PET) are used to evaluate the segmentation procedure in a setting of practical interest. Some data from a human brain imaging study with PET are used for further illustration.

ISSN:0162-1459    

DOI:10.1080/01621459.1997.10473606

出版商:Taylor & Francis Group    

年代:1997

数据来源: Taylor