In this paper, we consider the estimation of common breaks for linear panel data models by means screening and ranking algorithm. For static dynamic models, estimate regression coefficients using covariance generalized method moments, respectively, apply a algorithm on basis. The possible break points are first screened constructing local statistics based coefficient estimators, then further thresholding rule, finally final information criterion. Monte Carlo simulations demonstrate that...

This paper adopts a moving ratio statistic to monitor persistence change in long memory process.The limiting distribution of monitoring under the stationary null hypothesis is derived.We show that proposed scheme consistent for nonstationary change.In particular, sieve bootstrap approximation method proposed.The used determine critical values which depends on unknown parameter.The empirical size, power and average run length procedure are evaluated simulation study.Simulations indicate new...

In this paper, we propose a ratio test to detect the variance change in nonparametric regression models under both fixed and random design cases. The asymptotic validity of detection procedure is derived its finite sample performance evaluated via simulation study. Compared with existing cumulative sums (CUSUM) literature, our does not need estimate any scale parameter. particular, performs significantly better than those literature when shifts from large value small value. Finally,...

Abstract This article considers the sequential monitoring problem of variance change in stationary and non time series. We suggest a CUSUM squares procedure to detect infinite order moving average processes, residual autoregressive processes. Moreover, we introduce bandwidth parameter improve power when point does not occur at early stage monitoring. It is shown that both procedures have same null distribution. The are illustrated via simulation study an investigation daily Mexico/US...

In this article, we study the varying coefficient partially nonlinear model with measurement errors in nonparametric part. A local corrected profile least-square estimation procedure is proposed and asymptotic properties of resulting estimators are established. Further, a generalized likelihood ratio (GLR) statistic to test whether coefficients constant. The null distribution obtained residual-based bootstrap employed compute p-value statistic. Some simulations conducted evaluate performance...

Parameter estimation is an important component of statistical inference, and how to improve the accuracy parameter a key issue in research. This paper proposes linear Bayesian for estimating parameters misrecorded Poisson distribution. The method not only adopts prior information but also avoids cumbersome calculation posterior expectations. On premise ensuring stability computational results, we derived explicit solution estimation. Its superiority was verified through numerical simulations...

This article modifies the ratio test so that we can detect change which happens in latter half observations effectively. Moreover, propose a new on self-normalized numerator and denominator respectively. We establish asymptotic properties under null alternative hypotheses. also estimator to locate early or later change. check theoretical validity of block bootstrap approximation for modified tests give weak convergence proposed estimator. Simulations demonstrate procedures advantage An...

With the aim of providing better estimation for count data with overdispersion and/or excess zeros, we develop a novel method-optimal weighting based on cross-validation-for zero-inflated negative binomial model, where Poisson, binomial, and Poisson models are all included as its special cases. To facilitate selection optimal weight vector, K-fold cross-validation technique is adopted. Unlike jackknife model averaging discussed in Hansen Racine (2012), proposed method deletes one group...

The threshold autoregressive (TAR) model has received considerable attention in nonlinear time series literature. To weaken the impacts coming from uncertainty and to improve prediction accuracy, this paper develops a leave‐ ‐out forward‐validation averaging (LhoFVMA) method average predictions TAR model. We establish our method's asymptotic optimality sense of achieving lowest possible squared risk. Simulation experiments show that is generally more efficient than other methods. For...

This study considers the change-point test problem for time series based on self-normalization ratio statistic test, which is constructed using residuals obtained from a support vector regression (SVR)-autoregressive moving average (ARMA) model. Under null hypothesis, stationary process, and our converges to non-degenerate distribution. alternative there are change-points in series, diverges infinity. The simulations show that proposed new has better finite sample performance than other...

Multiple change-points estimation in panel data models is one of the popular topics statistics. In this article, we investigate multiple mean model based on a screening and ranking algorithm. Firstly, possible are initially screened local Secondly, threshold used to further screen change-points. Finally, final out using information criterion. Furthermore, consistency change-point estimators proved. The Monte Carlo simulation results show that proposed method can estimate number locations...

<abstract><p>In this paper, we propose a Dickey-Fuller difference statistic to sequentially detect the change-point that shift from an unit root process long-memory process. The limiting distribution of monitoring under null hypothesis as well its consistency alternative are proved. Simulations indicate new method can control empirical size even for heavy-tailed when using sieve bootstrap computing critical values. In particular, it performs significantly better than available in...