Control Variates for Monte Carlo Analysis of Nonlinear Statistical Models,I: Overview

Parameter values of nonlinear statistical models are typically estimated from data using iterative numerical procedures. The resulting joint sampling distribution of the parameter estimators is often intractable, resulting in the use of approximators or Monte Carlo simulation to determine properties of the sampling distribution.

This paper develops methods, using linear and higher-order approximators as control variates that reduce the variance of the Monte Carlo estimator by orders of magnitude. Estimation of means, higher-order raw moments, variances, covariances, and percentiles is considered.     

Sequential determination of the number of bootstrap samples

The primary purpose of this paper is that of developing a sequential Monte Carlo approximation to an ideal bootstrap estimate of the parameter of interest. Using the concept of fixed-precision approximation, we construct a sequential stopping rule for determining the number of bootstrap samples to be taken in order to achieve a specified precision of the Monte Carlo approximation. It is shown that the sequential Monte Carlo approximation is asymptotically efficient in the problems of estimation of the bias and standard error of a given statistic. Efficient bootstrap resampling is discussed and a numerical study is carried out for illustrating the obtained theoretical results.     

Estimation Under Purposive Sampling

Purposive sampling is described as a random selection of sampling units within the segment of the population with the most information on the characteristic of interest. Nonparametric bootstrap is proposed in estimating location parameters and the corresponding variances. An estimate of bias and a measure of variance of the point estimate are computed using the Monte Carlo method. The bootstrap estimator of the population mean is efficient and consistent in the homogeneous, heterogeneous, and two-segment populations simulated. The design-unbiased approximation of the standard error estimate differs substantially from the bootstrap estimate in severely heterogeneous and positively skewed populations.     

The effect of Monte Carlo approximation on coverage error of double-bootstrap confidence intervals

A double-bootstrap confidence interval must usually be approximated by a Monte Carlo simulation, consisting of two nested levels of bootstrap sampling. We provide an analysis of the coverage accuracy of the interval which takes account of both the inherent bootstrap and Monte Carlo errors. The analysis shows that, by a suitable choice of the number of resamples drawn at the inner level of bootstrap sampling, we can reduce the order of coverage error. We consider also the effects of performing a finite Monte Carlo simulation on the mean length and variability of length of two-sided intervals. An adaptive procedure is presented for the choice of the number of inner level resamples. The effectiveness of the procedure is illustrated through a small simulation study.     

Three Kolmogorov-Smirnov-type one-sample tests with improved power properties

Some simple point estimators are proposed for the three-parameter Weibull distribution. Both complete and type II censored sampling are considered. The biases and variances of these estimators are studied by Monte Carlo simulation.

Percentage points for the estimator of the shape parameter are also obtained by Monte Carlo simulation, which enables interval estimation and tests of hypotheses to be carried out for the shape parameter.     

Confidence intervals for the closed population size under inverse sampling without replacement

Inverse sampling is an appropriate design for the second phase of capture-recapture experiments which provides an exactly unbiased estimator of the population size. However, the sampling distribution of the resulting estimator tends to be highly right skewed for small recapture samples, so, the traditional Wald-type confidence intervals appear to be inappropriate. The objective of this paper is to study the performance of interval estimators for the population size under inverse recapture sampling without replacement. To this aim, we consider the Wald-type, the logarithmic transformation-based, the Wilson score, the likelihood ratio and the exact methods. Also, we propose some bootstrap confidence intervals for the population size, including the with-replacement bootstrap (BWR), the without replacement bootstrap (BWO), and the Rao–Wu’s rescaling method. A Monte Carlo simulation is employed to evaluate the performance of suggested methods in terms of the coverage probability, error rates and standardized average length. Our results show that the likelihood ratio and exact confidence intervals are preferred to other competitors, having the coverage probabilities close to the desired nominal level for any sample size, with more balanced error rate for exact method and shorter length for likelihood ratio method. It is notable that the BWO and Rao–Wu’s rescaling methods also may provide good intervals for some situations, however, those coverage probabilities are not invariant with respect to the population arguments, so one must be careful to use them.     

Bootstrap methods for heteroskedastic regression models: evidence on estimation and testing

This paper uses Monte Carlo simulation analysis to study the finite-sample behavior of bootstrap estimators and tests in the linear heteroskedastic model. We consider four different bootstrapping schemes, three of them specifically tailored to handle heteroskedasticity. Our results show that weighted bootstrap methods can be successfully used to estimate the variances of the least squares estimators of the linear parameters both under normality and under nonnormality. Simulation results are also given comparing the size and power of the bootstrapped Breusch-Pagan test with that of the original test and of Bartlett and Edgeworth-corrected tests. The bootstrap test was found to be robust against unfavorable regression designs.     

Bootstrap Procedures for Online Monitoring of Changes in Autoregressive Models

We compare the behavior of several bootstrap procedures for monitoring changes in the error distribution of autoregressive time series. The proposed procedures are designed to control the overall significance level and include classical tests based on the empirical distribution function as well as Fourier-type methods that utilize the empirical characteristic function, both functions being computed on the basis of properly estimated residuals. The Monte Carlo study incorporates different estimators and a variety of sampling situations with and without outliers.     

Bayesian analysis of complementary Poisson rate parameters with data subject to misclassification

We formulate closed-form Bayesian estimators for two complementary Poisson rate parameters using double sampling with data subject to misclassification and error free data. We also derive closed-form Bayesian estimators for two misclassification parameters in the modified Poisson model we assume. We use our results to determine credible sets for the rate and misclassification parameters. Additionally, we use MCMC methods to determine Bayesian estimators for three or more rate parameters and the misclassification parameters. We also perform a limited Monte Carlo simulation to examine the characteristics of these estimators. We demonstrate the efficacy of the new Bayesian estimators and highest posterior density regions with examples using two real data sets.     

Assessing circular error probable when the errors are elliptical normal

An important problem of continuing interest to engineers is the need to assess the circular error probable (CEP), a measure of the impact accuracy of a projectile or a measure of GPS point positioning accuracy. One of the challenges in addressing this problem is to construct some accurate confidence bounds or intervals for CEP in the small sample settings, where certain amount of systematic biases exist in testing experiments. Currently there is no general method available to deal with this challenge due to the intractability of the distributions of the existing CEP estimators. In this paper, in order to meet this challenge, several new approximate formulas are derived for calculating CEP, which are more accurate than the existing ones but still simple to use. Both exact and empirical expressions for the derivatives of CEP with respect to the population means and variances are also given. Using these formulas, three kinds of confidence bounds or intervals for CEP are proposed, which are based on the parametric bootstrap, the asymptotic distribution, and the Cornish–Fisher expansion, respectively. Moreover, a bias-corrected estimator of CEP is provided. The performances of these procedures are evaluated based on some Monte Carlo simulation studies. Both the theoretical and simulation results show that the Cornish–Fisher expansion-based procedure performs slightly better than the other two procedures when the downrange and cross-range variances are assumed the same. However, when these two variances are different, the simulation demonstrates that the bootstrap approach can be superior to the Cornish–Fisher for the small samples (say n=10), and vice versa for the moderate samples (say n=20).     

Error-rate estimation in two-group discriminant analysis using the linear discriminant function

Fisher's linear discriminant function, adapted by Anderson for allocating new observations into one of two existing groups, is considered in this paper. Methods of estimating the misclassification error rates are reviewed and evaluated by Monte Carlo simulations. The investigation is carried out under both ideal (Multivariate Normal data) and non-ideal (Multivariate Binary data) conditions. The assessment is based on the usual mean square error (MSE) criterion and also on a new criterion of optimism. The results show that although there is a common cluster of good estimators for both ideal and non-ideal conditions, the single best estimators vary with respect to the different criteria     

Importance sampling type estimators based on approximate marginal Markov chain Monte Carlo

We consider importance sampling (IS) type weighted estimators based on Markov chain Monte Carlo (MCMC) targeting an approximate marginal of the target distribution. In the context of Bayesian latent variable models, the MCMC typically operates on the hyperparameters, and the subsequent weighting may be based on IS or sequential Monte Carlo (SMC), but allows for multilevel techniques as well. The IS approach provides a natural alternative to delayed acceptance (DA) pseudo-marginal/particle MCMC, and has many advantages over DA, including a straightforward parallelization and additional flexibility in MCMC implementation. We detail minimal conditions which ensure strong consistency of the suggested estimators, and provide central limit theorems with expressions for asymptotic variances. We demonstrate how our method can make use of SMC in the state space models context, using Laplace approximations and time-discretized diffusions. Our experimental results are promising and show that the IS-type approach can provide substantial gains relative to an analogous DA scheme, and is often competitive even without parallelization.     

Comparison of bootstrap estimation intervals to forecast arithmetic mean and median air passenger demand

The aim of this paper is to compare passenger (pax) demand between airports based on the arithmetic mean (MPD) and the median pax demand (MePD). A three phases approach is applied. First phase, we use bootstrap procedures to estimate the distribution of the arithmetic MPD and the MePD for each block of routes distance; second phase, we use percentile, standard, bias corrected, and bias corrected accelerated methods to calculate bootstrap confidence bands for the MPD and the MePD; and third phase, we implement Monte Carlo (MC) experiments to analyse the finite sample performance of the applied bootstrap. Our results conclude that it is more meaningful to use the estimation of MePD rather than the estimation of MPD in the air transport industry. By carrying out MC experiments, we demonstrate that the bootstrap methods produce coverages close to the nominal for the MPD and the MePD.     

A fast Monte Carlo expectation–maximization algorithm for estimation in latent class model analysis with an application to assess diagnostic accuracy for cervical neoplasia in women with atypical glandular cells

In this article, we use a latent class model (LCM) with prevalence modeled as a function of covariates to assess diagnostic test accuracy in situations where the true disease status is not observed, but observations on three or more conditionally independent diagnostic tests are available. A fast Monte Carlo expectation–maximization (MCEM) algorithm with binary (disease) diagnostic data is implemented to estimate parameters of interest; namely, sensitivity, specificity, and prevalence of the disease as a function of covariates. To obtain standard errors for confidence interval construction of estimated parameters, the missing information principle is applied to adjust information matrix estimates. We compare the adjusted information matrix-based standard error estimates with the bootstrap standard error estimates both obtained using the fast MCEM algorithm through an extensive Monte Carlo study. Simulation demonstrates that the adjusted information matrix approach estimates the standard error similarly with the bootstrap methods under certain scenarios. The bootstrap percentile intervals have satisfactory coverage probabilities. We then apply the LCM analysis to a real data set of 122 subjects from a Gynecologic Oncology Group study of significant cervical lesion diagnosis in women with atypical glandular cells of undetermined significance to compare the diagnostic accuracy of a histology-based evaluation, a carbonic anhydrase-IX biomarker-based test and a human papillomavirus DNA test.     

Generalized Least Squares Model Averaging

In this article, we propose a method of averaging generalized least squares estimators for linear regression models with heteroskedastic errors. The averaging weights are chosen to minimize Mallows’ C_p-like criterion. We show that the weight vector selected by our method is optimal. It is also shown that this optimality holds even when the variances of the error terms are estimated and the feasible generalized least squares estimators are averaged. The variances can be estimated parametrically or nonparametrically. Monte Carlo simulation results are encouraging. An empirical example illustrates that the proposed method is useful for predicting a measure of firms’ performance.     

Empirical bayes estimation of several population means and variances under random sampling variances model

An extension of Kleffe–Rao model, an extended mixed model with random sampling variances, is considered. Empirical Bayes estimation is found to be very effective under such a model. The empirical Bayes estimators do not have a closed form. A second order Laplace approximation is proposed which works well for moderately large sample sizes. This approximation is specially useful when the uncertainties of the proposed empirical Bayes estimators are measured by the parametric bootstrap technique. A numerical example is considered to demonstrate the method.     

Posterior likelihood ratios for evaluation of forensic trace evidence given a two-level model on the data

In forensic science, in order to determine whether sets of traces are from the same source or not, it is widely advocated to evaluate evidential value of similarity of the traces by likelihood ratios (LRs). If traces are expressed by measurements following a two-level model with random effects and known variances, closed LR formulas are available given normality, or kernel density distributions, on the effects. For the known variances estimators are used though, which leads to uncertainty on the resulting LRs which is hard to quantify. The above is analyzed in an approach in which both effects and variances are random, following standard prior distributions on univariate data, leading to posterior LRs. For non-informative and conjugate priors, closed LR formulas are obtained that are interesting in structure and generalize a known result given fixed variance. A semi-conjugate prior on the model seems usable in many applications. It is described how to obtain credible intervals using Monte Carlo Markov Chain and regular simulation, and an example is described for comparison of XTC tablets based on MDMA content. In this way, uncertainty on LR estimation is expressed more clearly which makes the evidential value more transparent in a judicial context.     

New two-stage sampling designs based on neoteric ranked set sampling

Neoteric ranked set sampling (NRSS) is a recently developed sampling plan, derived from the well-known ranked set sampling (RSS) scheme. It has already been proved that NRSS provides more efficient estimators for population mean and variance compared to RSS and other sampling designs based on ranked sets. In this work, we propose and evaluate the performance of some two-stage sampling designs based on NRSS. Five different sampling schemes are proposed. Through an extensive Monte Carlo simulation study, we verified that all proposed sampling designs outperform RSS, NRSS, and the original double RSS design, producing estimators for the population mean with a lower mean square error. Furthermore, as with NRSS, two-stage NRSS estimators present some bias for asymmetric distributions. We complement the study with a discussion on the relative performance of the proposed estimators. Moreover, an additional simulation based on data of the diameter and height of pine trees is presented.     

Comparison of procedures for estimation of error rates in discriminant analysis under nonnormal population

The parametric and nonparametric methods for estimating the error rates in linear discriminant analysis are examined both in normal and in nonnormal situations. A Monte Carlo experiment was carried out under the assumption that two population distributions were characterized by a mixture of two multivariate normal distributions. The bootstrap bias-corrected apparent error rate compares favourably to other available estimators for nonnormal populations with small Mahalanobis distance. The methods for error estimation are also applied to a practical problem in medical diagnosis     

A location-invariant non-positive moment-type estimator of the extreme value index

This paper investigates a class of location invariant non-positive moment-type estimators of extreme value index, which is highly flexible due to the tuning parameter involved. Its asymptotic expansions and its optimal sample fraction in terms of minimal asymptotic mean square error are derived. A small scale Monte Carlo simulation turns out that the new estimators, with a suitable choice of the tuning parameter driven by the data itself, perform well compared to the known ones. Finally, the proposed estimators with a bootstrap optimal sample fraction are applied to an environmental data set.