scispace - formally typeset
Search or ask a question

Showing papers on "Semiparametric model published in 1993"


Book
01 Sep 1993
TL;DR: Asymptotic Inference for (Finite-Dimensional) Parametric Models as mentioned in this paper has been studied in the context of infinite-dimensional parametric models, where information bounds for Euclidean parameters in infinite-dimensional models have been derived.
Abstract: Introduction.- Asymptotic Inference for (Finite-Dimensional) Parametric Models.- Information Bounds for Euclidean Parameters in Infinite-Dimensional Models.- Euclidean Parameters: Further Examples.- Information Bounds for Infinite-Dimensional Parameters.- Infinite-Dimensional Parameters: Further Examples: Construction of Examples.

1,293 citations


Journal ArticleDOI
TL;DR: For the class of single-index models, this article constructed a semiparametric estimator of coefficients up to a multiplicative constant that exhibits 1 √ n -consistency and asymptotic normality.

1,208 citations


Journal ArticleDOI
TL;DR: In this paper, a binary response model of choice between automobile and transit for travel to work is estimated using fixed-and random-coefficients probit specifications, a semiparametric single-index specification, and a semi-parametric specification that permits arbitrary heteroskedasticity of unknown form.

153 citations


Journal ArticleDOI
TL;DR: In contrast, nonparametric regression estimation requires attention to (e.g., parameters and variables) but permits greatly reduced attention to the functional form of the regression model as mentioned in this paper.
Abstract: Current real estate statistical valuation involves the estimation of parameters within a posited specification. Suchparametric estimation requires judgment concerning model (1) variables; and (2) functional form. In contrast,nonparametric regression estimation requires attention to (1) but permits greatly reduced attention to (2). Parametric estimators functionally model the parameters and variables affectingE(y¦x) while nonparametric estimators directly modelpdf(y, x) and henceE(y¦x).

108 citations


Journal ArticleDOI
TL;DR: In this article, the authors provide a general framework for constructing specification tests for parametric and semiparametric models, and develop new specification tests using the general framework, which apply in time series and cross-sectional contexts.

97 citations


Journal ArticleDOI
Pedro Gozalo1
TL;DR: In this paper, a general framework for specification testing of the regression function in a nonparametric smoothing estimation context is proposed, which can be applied to cases as varied as testing for omission of variables, testing certain nonlinear restrictions in the regressors, and testing the correct specification of some parametric or semiparametric model of interest.
Abstract: This paper proposes a general framework for specification testing of the regression function in a nonparametric smoothing estimation context. The same analysis can be applied to cases as varied as testing for omission of variables, testing certain nonlinear restrictions in the regressors, and testing the correct specification of some parametric or semiparametric model of interest, for example, testing a certain type of nonlinearity of the regression function. Furthermore, the test can be applied to i.i.d. and time-series data, and some or all of the regressors are allowed to be discrete. A Monte Carlo simulation is used to assess the performance of the test in small and medium samples.

82 citations


Book ChapterDOI
TL;DR: In this paper, the authors present some of the results on semi-and non-parametric estimation of quantal response models, and discuss the consequences of adopting a misspecified parametric model.
Abstract: Publisher Summary This chapter presents some of the results on semi- and nonparametric estimation of quantal response models. It concentrates on methods for binary response models—that is, models in which the dependent variable has only two states, such as employed or unemployed. Semi- and nonparametric methods for binary response models are more highly developed than are such methods for models with multiple responses, and to date binary response is the only setting in which semi- or nonparametric methods have been applied. The chapter reviews parametric models, identifies the assumptions of these models that semi- and nonparametric methods relax, and explains the distinction between semi- and nonparametric methods. The consequences of adopting a misspecified parametric model are discussed as they are important because semi- and nonparametric methods are not needed if the errors resulting from use of a misspecified parametric model are small. The chapter discusses the problem of identifying behavioral parameters when parametric assumptions are relaxed, the rates of convergence and asymptotic efficiency in semiparametric models, the methods for semi- and nonparametric estimation of binary response models, the estimation from choice-based samples, the few applications of semiparametric estimators for binary response models, and models for multinomial responses.

76 citations


Journal ArticleDOI
TL;DR: In this article, the relative risk in a generalized Cox model with multivariate time dependent covariates is estimated using a penalized partial likelihood, and upper bounds on rate of convergence in a variety of norms are obtained.
Abstract: Nonparametric estimation of the relative risk in a generalized Cox model with multivariate time dependent covariates is considered. Estimation is based on a penalized partial likelihood. Using techniques from Andersen and Gill, and Cox and O'Sullivan, upper bounds on rate of convergence in a variety of norms are obtained. These upper bounds match the optimal rates available for linear nonparametric regression and density estimation. The results are uniform in the smoothing parameter, which is an important step for the analysis of data dependent rules for the selection of the smoothing parameter.

72 citations


Posted Content
TL;DR: In this article, the authors compare three types of nonparametric Maxfmum Zikelíhood estimators: Semt-nonparametrc MaxfmUM Zikellíhood, Semtnonparameter Zikeller estimators, and SingZe-equatton estimators.
Abstract: Esttmatfng a uxtge equatfon, account must be taken of the jact that ruages oj non-i,iorkers are not observed. For thts purpose, Heckman (1979) fntroduced the sample seLectfon modeL, consiating of troo equatíons: A(Línear) arage equatfon, explafning the potenttal Zog t,iage rate of every índívídual, incZudfng non-morkers, and a bfnary choice equatton, indícatíng ~uhether or not someone fs employed and the axige fa observed. TradíttonaL ML-eatfmatfon requfres a parametric specfftcatfon of the dtstributfon of the error terms, such as bivarfate normaZíty. Recently, a number of aemt-parametric estfmators have been developed mhích onZy requfre fndependence of the errors from the regressora ín both equations. We conatder three types o1 them: Semt-nonparametrtc Maxfmum Zikelíhood, ín mhich the parametera of interest and the densíty oJ the errors are eatimated afmuLtaneouaZy; 1}uo atage estfmators, generalizing the traditfonal Heckman ttoo step eatimator, míth semf-parametric estfmates o~ the bfnary chotce equatfon and a noriparametrfc correctton term added to the axige equatfon; and singZe-equatton estimators, neglecting irlJormation on non-raorkers. The estimators conaídered are asymptotícaZly normal, and aZZom Jor fnference. We preaent reaulta ,~or a sample of Dutch jemales and compare ~ith parametric !(L-estimatea. " We are grateful to the Netherlends Central Bureau of Statiatics (CBS) for providing the data. The views expressed i n this paper do not necessarily reflect the policies of the CBS. Financial support by the Netherlands Organisation of Scientific Research (NWO) and the Royal Netherlands Academy of Arts and Sciences (KNAW) is gratefully acknowledged by the first and second author, respectively. We are grateful for valuable commenta to Arie Kapteyn and seminar participants at CORE, (ironingen University, Tilburg University, and the Free University of Amsterdam.

22 citations


01 Jan 1993
TL;DR: In this paper, a semi-parametric approach to hazard rate estimation is proposed, which combines parametric and nonparametric features and uses a dynamic local likelihood approach to fit the locally most suitable member in a given parametric class of hazard rates.
Abstract: The best known methods for estimating hazard rate functions in survival analysis models are either purely parametric or purely nonparametric. The parametric ones are sometimes too biased while the nonparametric ones are sometimes too vari­ able. In the present paper a certain semiparametric approach to hazard rate estimation, proposed in Hjort (1991), is developed further, aiming to combine parametric and non­ parametric features. It uses a dynamic local likelihood approach to fit the locally most suitable member in a given parametric class of hazard rates, and amounts to a version of non parametric parameter smoothing within the parametric class. Thus the parametric hazard rate estimate at time 8 inserts a parameter estimate that also depends on 8. We study bias and variance properties of the resulting estimator and methods for choosing the local smoothing parameter. It is shown that dynamic likelihood estimation often leads to better performance than the purely nonparametric methods, while also having capacity for not losing much to the parametric methods in cases where the model being smoothed is adequate.

20 citations


Journal ArticleDOI
TL;DR: In this paper, an algorithm to derive the asymptotic lower bound for the information of the parameter governing the association between two survival times is presented. But the lower bound is not directly comparable to the upper bound in this paper.
Abstract: Clayton's model for association in bivariate survival data is both of intrinsic importance and an interesting example of a semiparametric estimation problem, that is, a problem where inference about a parameter is required in the presence of nuisance functions. The joint distribution of the two survival times in this model is absolutely continuous and a single parameter governs the association between the two survival times. In this paper we describe an algorithm to derive the asymptotic lower bound for the information of the parameter governing the association. We discuss the construction of one-step estimators and compare their performance to that of other estimators in a Monte Carlo study.

Journal ArticleDOI
TL;DR: In this paper, minimum distance estimates for the coefficient distributions in a general, semiparametric, random coefficient regression model were studied, and the analysis yields goodness-of-fit tests for the semi-parametric model, prediction regions for future responses, and confidence regions for the distribution of the random coefficients.
Abstract: Linear regression models with random coefficients express the idea that each individual sampled may have a different linear response function. Technically speaking, random coefficient regression encompasses a rich variety of submodels. These include deconvolution or affine-mixture models as well as certain classical linear regression models that have heteroscedastic errors, or errors-in-variables, or random effects. This paper studies minimum distance estimates for the coefficient distributions in a general, semiparametric, random coefficient regression model. The analysis yields goodness-of-fit tests for the semiparametric model, prediction regions for future responses, and confidence regions for the distribution of the random coefficients.


Journal ArticleDOI
TL;DR: In this article, a brief account of available FORTRAN Routines for computing nonparametric functional estimates, Frequently used in semiparametric problems, evaluated at each data point.
Abstract: The purpose of this note is to provide a brief account of available FORTRAN Routines for computing nonparametric functional estimates, Frequently used in semiparametric problems, evaluated at each data point. Then semiparametric estimates can be computed employing the use-favored economic software.

Journal ArticleDOI
TL;DR: In this article, it was shown that the estimator of panel data censored regression models proposed in Honore [21] is median unbiased when only one parameter is estimated, without parametric assumptions about the distribution of the error terms.
Abstract: This note proves that the estimator of panel data censored regression models proposed in Honore [21 is median unbiased when only one parameter is estimated. This result is obtained without parametric assumptions about the distribution of the error terms. The finite sample properties of estimators of semiparametric models are usually assessed via Monte Carlo simulations or by asymptotic expansions. These approaches often unveil regularities in a convincing way, but they share the weakness that the results are specific to the design under consideration. This note provides an example of an estimator of a semiparametric model that is median unbiased' in finite samples under (basically) the same conditions that are used to derive its asymptotic distribution. We consider the censored regression model with individual specific fixed effects studied in Honore [2] and we prove that the estimator proposed there is median unbiased in the special case where only one parameter is estimated. This result can be obtained even though the estimator is not symmetrically distributed. The model that we study is




Book ChapterDOI
01 Jan 1993
TL;DR: In this paper, the authors give an overview on several semiparametric methods used to model high-dimensional data, including generalized additive models (GAM), Alternating Conditional Expectations (ACE), Average Derivative Estimation (ADE), Semi-parametric Weighted Least Squares (SLES), SIM, Projection Pursuit Regression (PPR), and Sliced Inverse Regression(SIR).
Abstract: We give an overview on several semiparametric methods utilised to model high-dimensional data. Our approach is semiparametric in nature and is related to Generalised Linear Models. We focus on dynamic estimation techniques in this setting. In particular we discuss Generalized Additive Models (GAM), Alternating Conditional Expectations (ACE), Average Derivative Estimation (ADE), semiparametric weighted least squares (Single Index Models, SIM), Projection Pursuit Regression (PPR) , and Sliced Inverse Regression (SIR). Their performance in practice and theory is compared.

Journal ArticleDOI
TL;DR: In this paper, a semiparametric regression model was used to test the implications of the LCP hypothesis on the martingale property of consumption and several specification tests were performed on the U. S quarterly data from 1947 to 1990.
Abstract: This paper considers a semiparametric regression model to test the various implications of the Life Cycle-Permanent Income (LCP) hypothesis proposed by Hall (1978). The semiparametric regression model does not require any parametric assumption on the functional form of the unknown utility function in our analysis. In contrast, the linear regression models frequently used in the literature are justified under specific parametric forms of the utility function and may lead to a misleading conclusion on the LCP hypothesis if the parameterization is incorrect. Using both linear and semiparametric regression models, tests of the martingale property of consumption along with several specification tests are performed on the U. S quarterly data from 1947 to 1990. The results from the semiparametric model do not differ significantly from those from the linear model and suggest some evidences against the implications of the LCP hypothesis. [C14]