Keming Yu - Academia.edu (original) (raw)

Papers by Keming Yu

Electronic journal of statistics, 2024

Journal of Multivariate Analysis, 2020

IEEE Transactions on Reliability, 2014

Constant-stress procedures based on parametric lifetime distributions and models are often used f... more Constant-stress procedures based on parametric lifetime distributions and models are often used for accelerated life testing in product reliability experiments. Maximum likelihood estimation (MLE) is the typical statistical inference method. This paper presents a new inference method, named the random variable transformation (RVT) method, for Weibull constant-stress accelerated life tests with progressively Type-II right censoring (including ordinary Type-II right censoring). A two-parameter Weibull life distribution with a scale parameter that is a log-linear function of stress is used. RVT inference life distribution parameters and the log-linear function coefficients are provided. Exact confidence intervals for these parameters are also explored. Numerical comparisons of RVT-based estimates to MLE show that the proposed RVT inference is promising, in particular for small sample sizes.

Annals of the Institute of Statistical Mathematics, 2012

Estimating equation approaches have been widely used in statistics inference. Important examples ... more Estimating equation approaches have been widely used in statistics inference. Important examples of estimating equations are the likelihood equations. Since its introduction by Sir R. A. Fisher almost a century ago, maximum likelihood estimation (MLE) is still the most popular estimation method used for fitting probability distribution to data, including fitting lifetime distributions with censored data. However, MLE may produce substantial bias and even fail to obtain valid confidence intervals when data size is not large enough or there is censoring data. In this paper, based on nonlinear combinations of order statistics, we propose new estimation equation approaches for a class of probability distributions, which are particularly effective for skewed distributions with small sample sizes and censored data. The proposed approaches may possess a number of attractive properties such as consistency, sufficiency and uniqueness. Asymptotic normality of these new estimators is derived. The construction of new estimation equations and their numerical performance under different censored schemes are detailed via Weibull distribution and generalized exponential distribution.

Neurocomputing, 2021

Quantile regression has become a popular alternative to least squares regression for providing a ... more Quantile regression has become a popular alternative to least squares regression for providing a comprehensive description of the response distribution, and robustness against heavy-tailed error distributions. However, the nonsmooth quantile loss poses new challenges to distributed estimation in both computation and theoretical development. To address this challenge, we use a convolution-type smoothing approach and its Taylor expression to transform the nondifferentiable quantile loss function into a convex quadratic loss function, which admits a fast and scalable algorithm to perform optimization under massive and high-dimensional data. The proposed distributed estimators are both computationally and communication efficient. Moreover, only the gradient information is communicated at each iteration. Theoretically, we show that, after a certain number of iterations, the resulting estimator is statistically as efficient as the global estimator without any restriction on the number of machines. Both simulations and data analysis are conducted to illustrate the finite sample performance of the proposed methods.

Healthcare, 2020

This paper introduces several factors, namely, environmental pollution, medical level and environ... more This paper introduces several factors, namely, environmental pollution, medical level and environmental governance, into the Grossman’s production function for health. Then, an empirical analysis was conducted based on the 2004–2016 panel data of the city clusters along the middle reaches of the Yangtze River. Through the analysis, the author evaluated and compared how different factors affect the health of residents in the three city clusters: Changsha-Zhuzhou-Xiangtan (CZT) city cluster, Wuhan city cluster and circum-Poyang Lake (CPL) city cluster. The results show that: (1) In all three city clusters, economic growth can effectively improve the health of residents, and environmental pollution is also a key influencing factor of the health of residents. (2) Medical level has a close correlation with the health of residents. In the CZT city cluster, the medical level is positively correlated with the health of residents; in the CPL city cluster, the correlation is negative and take...

Journal of Statistical Planning and Inference, 1986

Some asymptotic results for a kernel type estimator of the quantile function from right-censored ... more Some asymptotic results for a kernel type estimator of the quantile function from right-censored data are obtained. The estimator is defined by n(p)h 1 Q(t)K((t-p)/hn)dt, which is smoother than the usual product-limit quantile function Q n(p)-inf{t: F n(t) k p), where F denotes the product-limit estimator of the lifetime distribution F. Under the random censorship model and general conditions on hn ,K, and F 0 , the asymptotic normality of Qn(p) is proven. In addition, an approximation to n is shown to be asymptotically uniformly equivalent to Qn in mean square.

Journal of Statistical Computation and Simulation, 2013

This article may be used for research, teaching, and private study purposes. Any substantial or s... more This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Journal of Multivariate Analysis, 2007

We present methods to handle error-in-variables models. Kernel-based likelihood score estimating ... more We present methods to handle error-in-variables models. Kernel-based likelihood score estimating equation methods are developed for estimating conditional density parameters. In particular, a semiparametric likelihood method is proposed for sufficiently using the information in the data. The asymptotic distribution theory is derived. Small sample simulations and a real data set are used to illustrate the proposed estimation methods.

International Statistical Review, 2001

Folklore has it that a very simple supervised classification rule, based on the typically false a... more Folklore has it that a very simple supervised classification rule, based on the typically false assumption that the predictor variables are independent, can be highly effective, and often more effective than sophisticated rules. We examine the evidence For this, both empirical, as observed in real data applications, and theoretical, summarising explanations for why this simple rule might be effective.

Computational Statistics, 2012

We propose a model for multinomial probit factor analysis by assuming t-distribution error in pro... more We propose a model for multinomial probit factor analysis by assuming t-distribution error in probit factor analysis. To obtain maximum likelihood estimation, we use the Monte Carlo expectation maximization algorithm with its M-step greatly simplified under conditional maximization and its E-step made feasible by Monte Carlo simulation. Standard errors are calculated by using Louis's method. The methodology is illustrated with numerical simulations.

A new multivariate random walk model with slowly changing drift and cross-correlations for multiv... more A new multivariate random walk model with slowly changing drift and cross-correlations for multivariate processes is introduced and investigated in detail. In the model, not only the drifts and the cross-covariances but also the cross-correlations between single series are allowed to change slowly over time. The model can accompany any number of components such as many number of assets. The model is particularly useful for modelling and forecasting the value of financial portfolios under very complex market conditions. Kernel estimation of local covariance matrix is used. The integrated effect of the estimation errors involved in estimating the integrated processes is derived. Practical relevance of the model and estimation is illustrated by application to several foreign exchange rates.

A new multivariate random walk model with slowly changing parameters is introduced and investigat... more A new multivariate random walk model with slowly changing parameters is introduced and investigated in detail. Nonparametric estimation of local covariance matrix is proposed. The asymptotic distributions, including asymptotic biases, variances and covariances of the proposed estimators are obtained. The properties of the estimated value of a weighted sum of individual nonparametric estimators are also studied in detail. The integrated effect of the estimation errors from the estimation for the difference series to the integrated processes is derived. Practical relevance of the model and estimation is illustrated by application to several foreign exchange rates.

Applied Stochastic Models in Business and Industry, 2016

The parametric conditional autoregressive expectiles (CARE) models have been developed by [1] to ... more The parametric conditional autoregressive expectiles (CARE) models have been developed by [1] to estimate expectiles that can be used to assess value at risk (VaR) and expected shortfall (ES). The challenge lies in parametric CARE modeling is specification of a parametric form. To avoid any model misspecification, we propose a nonparametric CARE model via neural network. The nonparametric CARE model can be estimated by a classical gradient based nonlinear optimization algorithm. We then apply the nonparametric CARE model to estimating VaR and ES of six stock indices. Empirical results for the new model is competitive with those classical models and parametric CARE models.

Axioms, 2022

Functional data, which provides information about curves, surfaces or anything else varying over ... more Functional data, which provides information about curves, surfaces or anything else varying over a continuum, has become a commonly encountered type of data. The k-nearest neighbor (kNN) method, as a nonparametric method, has become one of the most popular supervised machine learning algorithms used to solve both classification and regression problems. This paper is devoted to the k-nearest neighbor (kNN) estimators of the nonparametric functional regression model when the observed variables take values from negatively associated (NA) sequences. The consistent and complete convergence rate for the proposed kNN estimator is first provided. Then, numerical assessments, including simulation study and real data analysis, are conducted to evaluate the performance of the proposed method and compare it with the standard nonparametric kernel approach.

In this thesis, attention will be mainly focused on the local linear kernel regression quantile e... more In this thesis, attention will be mainly focused on the local linear kernel regression quantile estimation. Different estimators within this class have been proposed, developed asymptotically and applied to real applications. I include algorithmdesign and selection of smoothing parameters. Chapter 2 studies two estimators, first a single-kernel estimator based on "check function" and a bandwidth selection rule is proposed based on the asymptotic MSE of this estimator. Second a recursive double-kernel estimator which extends Fan et al's (1996) density estimator, and two algorithms are given for bandwidth selection. In Chapter 3, a comparison is carried out of local constant fitting and local linear fitting using MSEs of the estimates as a criterion. Chapter 4 gives a theoretical summary and a simulation study of local linear kernel estimation of conditional distribution function. This has a special interest in itself as well as being related to regression quantiles. In ...

Lattice Boltzmann methods are kinetic descriptions of fluid flow that are efficiently implemented... more Lattice Boltzmann methods are kinetic descriptions of fluid flow that are efficiently implemented through a stream and collide approach. The collision operation is typically an approximation of the microscopic physics, with the BGK linear approximation a widely used choice. It has a wide range of application in computational fluid dynamics. The honor thesis starts with an introduction to the kinetic theory of gas. First, we introduce the Liouville Equation that describes the evolution of particle distribution function of a system in phase space. With an analysis of the Liouville Equation, BBGKY hierarchy is introduced for the reduced distribution evolution. Using the Bogoliubov Hypothesis, we are able to close the hierarchy equation and derive the Boltzmann's Equation for kinetic theory of gas. An analysis of Boltzmann's Equation is presented, with the center of analysis being the Chapman-Enskog analysis that reproduces the Navier-Stokes Equation in its expansion. Then, the ...

Like mean, quantile and variance, mode is also an important measure of central tendency and data ... more Like mean, quantile and variance, mode is also an important measure of central tendency and data summary. Many practical questions often focus on "Which element (gene or file or signal) occurs most often or is the most typical among all elements in a network?". In such cases mode regression provides a convenient summary of how the regressors affect the conditional mode and is totally different from other regression models based on conditional mean or conditional quantile or conditional variance. Some inference methods have been used for mode regression but none of them from the Bayesian perspective. This paper introduces Bayesian mode regression by exploring three different approaches. We start from a parametric Bayesian model by employing a likelihood function that is based on a mode uniform distribution. It is shown that irrespective of the original distribution of the data, the use of this special uniform distribution is a very natural and effective way for Bayesian mode regression. Posterior estimates based on this parametric likelihood, even under misspecification, are consistent and asymptotically normal. We then develop a nonparametric Bayesian model by using Dirichlet process (DP) mixtures of mode uniform distributions and finally we explore Bayesian empirical likelihood mode regression by taking empirical likelihood into a Bayesian framework. The paper also demonstrates that a variety of improper priors for the unknown model parameters yield a proper joint posterior. The proposed approach is illustrated using simulated datasets and a real data set.

Computational Optimization in Engineering - Paradigms and Applications, 2017

Journal of Inequalities and Applications, 2018

In the present article, by utilizing some inequalities for linearly negative quadrant dependent r... more In the present article, by utilizing some inequalities for linearly negative quadrant dependent random variables, we discuss the uniformly asymptotic normality of sample quantiles for linearly negative quadrant dependent samples under mild conditions. The rate of uniform asymptotic normality is presented and the rate of convergence is near O(n-1/4 log n) when the third moment is finite, which extends and improves the corresponding results of Yang et al.