Self-weighted quasi-maximum exponential likelihood estimator for ARFIMA-GARCH models

In this paper, a local self-weighted quasi-maximum exponential likelihood estimator for ARFIMA-GARCH models is proposed, asymptotic normality of this estimator is derived under the existence of second moment including stationary and non-stationary cases. A simulation study is given to evaluate the performance of the proposed self-weighted QMELE under the stationary case.     

Bayesian analysis of the linear regression model with an edgeworth series prior distribution

The purpose of the present investigation 1s to observe the effect of departure from normahty of the prior distribution of regresslon parameters on the Bayman analysis of a h e a r regresslon model Assuming an Edgeworth serles prior distribution for the regresslon coefficients and gamma prior for the disturbances precision, the expressions for the posterlor distribution, posterlor mean and Bayes risk under a quadratic loss function are obtalned The results of a numerical evaluation are also analyzed     

On the restricted almost unbiased estimators in linear regression

In this paper, the restricted almost unbiased ridge regression estimator and restricted almost unbiased Liu estimator are introduced for the vector of parameters in a multiple linear regression model with linear restrictions. The bias, variance matrices and mean square error (MSE) of the proposed estimators are derived and compared. It is shown that the proposed estimators will have smaller quadratic bias but larger variance than the corresponding competitors in literatures. However, they will respectively outperform the latter according to the MSE criterion under certain conditions. Finally, a simulation study and a numerical example are given to illustrate some of the theoretical results.     

Some asymptotic results in finite populations

In this paper we investigate limiting properties of predictors of some finite population quantities. Both, the sample size and the population size are considered to become large. Limiting properties like consistency and asymptotic normality of the best linear unbiased predictors of the population total and of the finite population regression coefficient are investigated.     

ESTIMATION PROCEDURES FOR CATEGORICAL SURVEY DATA WITH NONIGNORABLE NONRESPONSE

We consider surveys with one or more callbacks and use a series of logistic regressions to model the probabilities of nonresponse at first contact and subsequent callbacks. These probabilities are allowed to depend on covariates as well as the categorical variable of interest and so the nonresponse mechanism is nonignorable. Explicit formulae for the score functions and information matrices are given for some important special cases to facilitate implementation of the method of scoring for obtaining maximum likelihood estimates of the model parameters. For estimating finite population quantities, we suggest the imputation and prediction approaches as alternatives to weighting adjustment. Simulation results suggest that the proposed methods work well in reducing the bias due to nonresponse. In our study, the imputation and prediction approaches perform better than weighting adjustment and they continue to perform quite well in simulations involving misspecified response models.     

Non Parametric Regression Quantile Estimation for Dependent Functional Data under Random Censorship: Asymptotic Normality

In this paper we study a smooth estimator of the regression quantile function in the censorship model when the covariates take values in some abstract function space. The main goal of this paper is to establish the asymptotic normality of the kernel estimator of the regression quantile, under α-mixing condition and, on the concentration properties on small balls probability measure of the functional regressors. Some applications and particular cases are studied. This study can be applied in time series analysis to the prediction and building confidence bands. Some simulations are drawn to lend further support to our theoretical results and to compare the quality of behavior of the estimator for finite samples with different rates of censoring and sizes.     

Rejoinder: "Comment on 'a new Statistic for Testing Suspected Outliers'"

The values of the power of Tiku's (1975) T statistic for testing outliers in normal samples are evaluated. The statistic T is shown to be more powerful than other comparable statistics under Tiku's outlier model, although slightly less powerful under Dixon's (1950) contamination model.     

Nonparametric Estimation for FBSDEs Models with Applications in Finance

As a continuous-time model, forward-backward stochastic differential equations (in short FBSDEs) have been successfully applied in mathematical finance, e.g., European option pricing for either a small or a large investor in a Markovian market. However, the correct FBSDEs model for a specific topic can neither be provided automatically by financial market nor derived from theory of mathematical finance. In this article, a nonparametric FBSDEs model is adopted for its flexibility and robustness, and the estimators of the functional coefficients of the FBSDEs model are obtained. The asymptotic properties of the estimators are also discussed. A simulation is performed to test the feasibility of our method.     

A decision theoretic approach to ridits

Since its inception, ridit analyses has been in widespread use in epidemic-logic studies where the data are ordered but are not on an interval scale. However, no mathematical properties of ridits have been given. In this paper, we use a squared error loss function to show that, for a particular class of distribution functions, ridits form a best invariant estimate of the unknown distribution function. Under another class of distribution functions, we derive another estimate, m-ridits, of the distribution function. Data are used to compare these two scores with the scores obtained from the empirical distribution function and the original scores used on the data. The results indicate that, although these scores are numerically different, the same inferences can be drawn.     

Homogeneity Tests for Related Samples Under Censoring

We propose nonparametric homogeneity tests for related samples against much wider than location (or scale) class of alternatives including possible crossings of marginal cumulative distribution functions. The tests can be used in the case of complete and censored samples. Asymptotic distribution of the test statistics is investigated.