Semiparametric Likelihood Research Articles

Semi-parametric Dynamic Models for Longitudinal Ordinal Categorical Data

The over all regression function in a semi-parametric model involves a partly specified regression function in some primary covariates and a non-parametric function in some other secondary covariates. This type of semi-parametric models in a longitudinal setup has recently been discussed extensively both for repeated Poisson and negative binomial count data. However, when it is appropriate to interpret the longitudinal binary responses through a binary dynamic logits model, the inferences for semi-parametric Poisson and negative binomial models cannot be applied to such binary models as these models unlike the count data models produce recursive means and variances containing the dynamic dependence or correlation parameters. In this paper, we consider a general multinomial dynamic logits model in a semi-parametric setup first to analyze nominal categorical data in a semi-parametric longitudinal setup, and then modify this model to analyze ordinal categorical data. The ordinal responses are fitted by using a cumulative semi-parametric multinomial dynamic logits model. For the benefits of practitioners, a step by step estimation approach is developed for the non-parametric function, and for both regression and dynamic dependence parameters. In summary, a kernel-based semi-parametric weighted likelihood approach is used for the estimation of the non-parametric function. This weighted likelihood estimate for the non-parametric function is shown to be consistent. The regression and dynamic dependence parameters of the model are estimated by maximizing an approximate semi-parametric likelihood function for the parameters, which is constructed by replacing the non-parametric function with its consistent estimate. Asymptotic properties including the proofs for the consistency of the likelihood estimators of the regression and dynamic dependence parameters are discussed.

Semiparametric Inference in a Genetic Mixture Model

ABSTRACTIn genetic backcross studies, data are often collected from complex mixtures of distributions with known mixing proportions. Previous approaches to the inference of these genetic mixture models involve parameterizing the component distributions. However, model misspecification of any form is expected to have detrimental effects. We propose a semiparametric likelihood method for genetic mixture models: the empirical likelihood under the exponential tilting model assumption, in which the log ratio of the probability (density) functions from the components is linear in the observations. An application to mice cancer genetics involves random numbers of offspring within a litter. In other words, the cluster size is a random variable. We wish to test the null hypothesis that there is no difference between the two components in the mixture model, but unfortunately we find that the Fisher information is degenerate. As a consequence, the conventional two-term expansion in the likelihood ratio statistic does not work. By using a higher-order expansion, we are able to establish a nonstandard convergence rate N− 1/4 for the odds ratio parameter estimator . Moreover, the limiting distribution of the empirical likelihood ratio statistic is derived. The underlying distribution function of each component can also be estimated semiparametrically. Analogously to the full parametric approach, we develop an expectation and maximization algorithm for finding the semiparametric maximum likelihood estimator. Simulation results and a real cancer application indicate that the proposed semiparametric method works much better than parametric methods. Supplementary materials for this article are available online.

Case-cohort studies with interval-censored failure time data.

The case-cohort design has been widely used as a means of cost reduction in assembling or measuring expensive covariates in large cohort studies. The existing literature on the case-cohort design is mainly focused on right-censored data. In practice, however, the failure time is often subject to interval-censoring; it is known only to fall within some random time interval. In this paper, we consider the case-cohort study design for interval-censored failure time and develop a sieve semiparametric likelihood approach for analyzing data from this design under the proportional hazards model. We construct the likelihood function using inverse probability weighting and build the sieves with Bernstein polynomials. The consistency and asymptotic normality of the resulting regression parameter estimator are established and a weighted bootstrap procedure is considered for variance estimation. Simulations show that the proposed method works well for practical situations, and an application to real data is provided.

Efficient inverse probability weighting method for quantile regression with nonignorable missing data

ABSTRACTQuantitle regression (QR) is a popular approach to estimate functional relations between variables for all portions of a probability distribution. Parameter estimation in QR with missing data is one of the most challenging issues in statistics. Regression quantiles can be substantially biased when observations are subject to missingness. We study several inverse probability weighting (IPW) estimators for parameters in QR when covariates or responses are subject to missing not at random. Maximum likelihood and semiparametric likelihood methods are employed to estimate the respondent probability function. To achieve nice efficiency properties, we develop an empirical likelihood (EL) approach to QR with the auxiliary information from the calibration constraints. The proposed methods are less sensitive to misspecified missing mechanisms. Asymptotic properties of the proposed IPW estimators are shown under general settings. The efficiency gain of EL-based IPW estimator is quantified theoretically. Simulation studies and a data set on the work limitation of injured workers from Canada are used to illustrated our proposed methodologies.

Orthogonality of the mean and error distribution in generalized linear models

ABSTRACTWe show that the mean-model parameter is always orthogonal to the error distribution in generalized linear models. Thus, the maximum likelihood estimator of the mean-model parameter will be asymptotically efficient regardless of whether the error distribution is known completely, known up to a finite vector of parameters, or left completely unspecified, in which case the likelihood is taken to be an appropriate semiparametric likelihood. Moreover, the maximum likelihood estimator of the mean-model parameter will be asymptotically independent of the maximum likelihood estimator of the error distribution. This generalizes some well-known results for the special cases of normal, gamma, and multinomial regression models, and, perhaps more interestingly, suggests that asymptotically efficient estimation and inferences can always be obtained if the error distribution is non parametrically estimated along with the mean. In contrast, estimation and inferences using misspecified error distributions or variance functions are generally not efficient.

A new approach to regression analysis of censored competing-risks data.

An approximate likelihood approach is developed for regression analysis of censored competing-risks data. This approach models directly the cumulative incidence function, instead of the cause-specific hazard function, in terms of explanatory covariates under a proportional subdistribution hazards assumption. It uses a self-consistent iterative procedure to maximize an approximate semiparametric likelihood function, leading to an asymptotically normal and efficient estimator of the vector of regression parameters. Simulation studies demonstrate its advantages over previous methods.

Fitting Cox Models with Doubly Censored Data Using Spline-Based Sieve Marginal Likelihood.

In some applications, the failure time of interest is the time from an originating event to a failure event, while both event times are interval censored. We propose fitting Cox proportional hazards models to this type of data using a spline-based sieve maximum marginal likelihood, where the time to the originating event is integrated out in the empirical likelihood function of the failure time of interest. This greatly reduces the complexity of the objective function compared with the fully semiparametric likelihood. The dependence of the time of interest on time to the originating event is induced by including the latter as a covariate in the proportional hazards model for the failure time of interest. The use of splines results in a higher rate of convergence of the estimator of the baseline hazard function compared with the usual nonparametric estimator. The computation of the estimator is facilitated by a multiple imputation approach. Asymptotic theory is established and a simulation study is conducted to assess its finite sample performance. It is also applied to analyzing a real data set on AIDS incubation time.

Likelihood Inferences on Semiparametric Odds Ratio Model

A flexible semiparametric odds ratio model has been proposed to unify and to extend both the log-linear model and the joint normal model for data with a mix of discrete and continuous variables. The semiparametric odds ratio model is particularly useful for analyzing biased sampling designs. However, statistical inference of the model has not been systematically studied when more than one nonparametric component is involved in the model. In this article, we study the maximum semiparametric likelihood approach to estimation and inference of the semiparametric odds ratio model. We show that the maximum semiparametric likelihood estimator of the odds ratio parameter is consistent and asymptotically normally distributed. We also establish statistical inference under a misspecified semiparametric odds ratio model, which is important when handling weak identifiability in conditionally specified models under biased sampling designs. We use simulation studies to demonstrate that the proposed approaches have satisfactory finite sample performance. Finally, we illustrate the proposed approach by analyzing multiple traits in a genome-wide association study of high-density lipid protein. Supplementary materials for this article are available online.

Spline-based semiparametric estimation of a zero-inflated Poisson regression single-index model

When the number of zeros in a count dataset exceeds the accommodation of the probability mass of a regular Poisson distribution at zero, the zero-inflated Poisson (ZIP) distribution is often used. To characterize the potential non-linear effects of covariates and avoid the “curse of dimensionality”, we propose a spline-based ZIP regression single-index model. B-splines are employed to estimate the unknown smooth function. A modified Fisher scoring method is proposed to simultaneously estimate the linear coefficients and the regression function. It is shown that the spline estimator of the nonparametric component is uniformly consistent, and achieves the optimal convergence rate under the smooth condition, and that the estimators of regression parameters are asymptotically normal and efficient. The spline-based semiparametric likelihood ratio test is also established. Moreover, a direct and consistent variance estimation method based on least-squares estimation is proposed. Simulations are performed to evaluate the proposed method.

Semiparametric score driven volatility models

A new semiparametric observation-driven volatility model is proposed. In contrast to the standard semiparametric generalized autoregressive conditional heteroskedasticity (GARCH) model, the form of the error density has a direct influence on both the semiparametric likelihood and the volatility dynamics. The estimator is shown to consistently estimate the conditional pseudo true parameters of the model. Simulation-based evidence and an empirical application to stock return data confirm that the new statistical model realizes substantial improvements compared to GARCH type models and quasi-maximum likelihood estimation if errors are fat-tailed and possibly skewed.

Semiparametric likelihood inference for left-truncated and right-censored data.

This paper proposes a new estimation procedure for the survival time distribution with left-truncated and right-censored data, where the distribution of the truncation time is known up to a finite-dimensional parameter vector. The paper expands on the Vardis multiplicative censoring model (Vardi, 1989. Multiplicative censoring, renewal processes, deconvolution and decreasing density: non-parametric estimation. Biometrika 76: , 751-761), establishes the connection between the likelihood under a generalized multiplicative censoring model and that for left-truncated and right-censored survival time data, and derives an Expectation-Maximization algorithm for model estimation. A formal test for checking the truncation time distribution is constructed based on the semiparametric likelihood ratio test statistic. In particular, testing the stationarity assumption that the underlying truncation time is uniformly distributed is performed by embedding the null uniform truncation time distribution in a smooth alternative (Neyman, 1937. Smooth test for goodness of fit. Skandinavisk Aktuarietidskrift 20: , 150-199). Asymptotic properties of the proposed estimator are established. Simulations are performed to evaluate the finite-sample performance of the proposed methods. The methods and theories are illustrated by analyzing the Canadian Study of Health and Aging and the Channing House data, where the stationarity assumption with respect to disease incidence holds for the former but not the latter.

An extended hazard model with longitudinal covariates

In clinical trials and other medical studies, it has become increasingly common to observe simultaneously an event time of interest and longitudinal covariates. In the literature, joint modelling approaches have been employed to analyse both survival and longitudinal processes and to investigate their association. However, these approaches focus mostly on developing adaptive and flexible longitudinal processes based on a prespecified survival model, most commonly the Cox proportional hazards model. In this paper, we propose a general class of semiparametric hazard regression models, referred to as the extended hazard model, for the survival component. This class includes two popular survival models, the Cox proportional hazards model and the accelerated failure time model, as special cases. The proposed model is flexible for modelling event data, and its nested structure facilitates model selection for the survival component through likelihood ratio tests. A pseudo joint likelihood approach is proposed for estimating the unknown parameters and components via a Monte Carlo em algorithm. Asymptotic theory for the estimators is developed together with theory for the semiparametric likelihood ratio tests. The performance of the procedure is demonstrated through simulation studies. A case study featuring data from a Taiwanese HIV/AIDS cohort study further illustrates the usefulness of the extended hazard model.

Semi-Parametric Maximum Likelihood Method for Interaction in Case-Mother Control-Mother Designs: PackageSPmlficmcm

The analysis of interaction effects involving genetic variants and environmental exposures on the risk of adverse obstetric and early-life outcomes is generally performed using standard logistic regression in the case-mother and control-mother design. However such an analysis is inefficient because it does not take into account the natural family-based constraints present in the parent-child relationship. Recently, a new approach based on semi-parametric maximum likelihood estimation was proposed. The advantage of this approach is that it takes into account the parental relationship between the mother and her child in estimation. But a package implementing this method has not been widely available. In this paper, we present SPmlficmcm, an R package implementing this new method and we propose an extension of the method to handle missing offspring genotype data by maximum likelihood estimation. Our choice to treat missing data of the offspring genotype was motivated by the fact that in genetic association studies where the genetic data of mother and child are available, there are usually more missing data on the genotype of the offspring than that of the mother. The package builds a non-linear system from the data and solves and computes the estimates from the gradient and the Hessian matrix of the log profile semi-parametric likelihood function. Finally, we analyze a simulated dataset to show the usefulness of the package.

Likelihood-based methods for evaluating principal surrogacy in augmented vaccine trials.

There is growing interest in assessing immune biomarkers, which are quick to measure and potentially predictive of long-term efficacy, as surrogate endpoints in randomized, placebo-controlled vaccine trials. This can be done under a principal stratification approach, with principal strata defined using a subject's potential immune responses to vaccine and placebo (the latter may be assumed to be zero). In this context, principal surrogacy refers to the extent to which vaccine efficacy varies across principal strata. Because a placebo recipient's potential immune response to vaccine is unobserved in a standard vaccine trial, augmented vaccine trials have been proposed to produce the information needed to evaluate principal surrogacy. This article reviews existing methods based on an estimated likelihood and a pseudo-score (PS) and proposes two new methods based on a semiparametric likelihood (SL) and a pseudo-likelihood (PL), for analyzing augmented vaccine trials. Unlike the PS method, the SL method does not require a model for missingness, which can be advantageous when immune response data are missing by happenstance. The SL method is shown to be asymptotically efficient, and it performs similarly to the PS and PL methods in simulation experiments. The PL method appears to have a computational advantage over the PS and SL methods.

Spline estimation of generalised monotonic regression

We develop a simple and practical, yet flexible spline estimation method for semiparametric generalised linear models with monotonicity constraints. We propose to approximate the unknown monotone function by monotone B-splines, and employ generalised Rosen algorithm to compute the estimates. We show that the spline estimate of the nonparametric component achieves the optimal rate of convergence under the smooth condition, and that the estimates of regression parameters are asymptotically normal and efficient. The spline-based semiparametric likelihood ratio test (LRT) is also established. Moreover, a direct variance estimation method based on least-squares estimation is proposed. The finite sample performance of the spline estimates is evaluated by a Monte Carlo study. The methodology is illustrated on an air pollution study.

Semiparametric maximum likelihood inference by using failed contact attempts to adjust for nonignorable nonresponse

In marketing research, social science and epidemiological studies, call-back of nonrespondents is standard. If respondents and nonrespondents tend to give different answers, the missing data are called nonignorable, and using them alone may produce biased results. To extend earlier work on nonresponse in the presence of call-backs, Alho (1990) proposed modelling the probability of response at each attempt through logistic regression, where outcomes of interest and covariates are explanatory variables. In this paper we propose a semiparametric maximum likelihood approach, and discuss large-sample properties and the semiparametric likelihood ratio statistic used to test whether the data are missing completely at random. Simulations are conducted to evaluate this approach and a modification of the method of Alho (1990). Data from the National Health Interview Survey are used for illustration.

Semi-parametric Efficient Inference for Heteroscedastic Semivarying-coefficient Models

Semivarying-coefficient models with heteroscedastic errors are frequently used in statistical modeling. When the error is conditional heteroskedastic, Ahmad, et al. (2005) proposed a general series method to obtain an efficient estimation. In this article we study the heteroscedastic semi-varying coefficient models with a nonparametric variance function, not only use the semi-parametric efficient normal approximation method to derive a family of semi-parametric efficient estimator, but also use the semi-parametric efficient empirical likelihood method to construct the efficient empirical likelihood confidence regions. The proposed estimators retain the double robustness feature of semi-parametric efficient estimator.

Semiparametric empirical likelihood estimation for two-stage outcome-dependent sampling under the frame of generalized linear models

Epidemiologic studies use outcome-dependent sampling (ODS) schemes where, in addition to a simple random sample, there are also a number of supplement samples that are collected based on outcome variable. ODS scheme is a cost-effective way to improve study efficiency. We develop a maximum semiparametric empirical likelihood estimation (MSELE) for data from a two-stage ODS scheme under the assumption that given covariate, the outcome follows a general linear model. The information of both validation samples and nonvalidation samples are used. What is more, we prove the asymptotic properties of the proposed MSELE.

Semiparametric empirical likelihood confidence intervals for the difference of areas under two correlated ROC curves under density ratio model

The area under the receiver operating characteristic (ROC) curve, AUC, is commonly used to assess the ability of a diagnostic test to correctly classify individuals into diseased and nondiseased populations. When there are two diagnostic tests available, it is of interest to evaluate and compare their performances. Based on the difference of two placement values, we propose a two-sample empirical likelihood method for comparing AUCs of two ROC curves. The proposed empirical likelihood ratio statistic converges in distribution to a scaled chi-squared random variable. Simulation results show that the proposed empirical likelihood method has better finite-sample performance than other competitors.

Semiparametric empirical likelihood confidence intervals for AUC under a density ratio model

Inferences on the area under a receiver operating characteristic curve (AUC) are usually based on a fully parametric approach or a fully nonparametric approach. A semiparametric empirical likelihood method is proposed to construct confidence intervals for AUC by assuming a density ratio model for the diseased and non-diseased population densities. The limiting distribution of the semiparametric empirical log likelihood ratio statistic for AUC has a scaled chi-square distribution. The proposed semiparametric empirical likelihood approach is shown, via a simulation study, to be more robust than a fully parametric approach and is more accurate than a fully nonparametric approach. Some results on simulation and an analysis of two real examples are presented.