mk体育
2019年统计学前沿学术研讨会
暨博士后论坛
会议手册
河南 开封
7月21日-7月23日
会议注意事项 |
1.从开封北站打车到mk体育金明校区大西门旁中州颐和酒店大概15分钟，费用约15元 |
2.从宋城路站打车到mk体育新校区大西门旁中州颐和酒店大概15分钟，费用约13元 |
3.从新郑机场到开封每隔一个小时一趟城际高铁，大约28分钟，到开封北站或者宋城路站均可 |
4.从酒店走到mk体育官网大约600米，约5分钟，路线见百度地图截图
|
5.会议用餐: 早餐在中州颐和酒店一楼， 21日晚餐及22，23日午餐中州颐和酒店酒店一楼， 22日晚餐中州颐和酒店二楼。 |
Schedule |
2019年7月22日，星期一上午，数学院一楼报告厅 |
时间 |
报告人 |
报告题目 |
8:00-8:20 |
开幕式 |
主持人：冯淑霞 |
公司领导致辞 |
公司一楼照相 |
主持人：薛留根（北京工业大学） |
8:20-8:50 |
梁汉营 （同济大学） |
Quantile regression for partially linear varying-coefficient model with censoring indicators missing at random |
8:50-9:20 |
吕绍高 （南京审计大学） |
Asymptotic efficiency of imputation based semi-supervised learning in misspecified setting |
9:20-9:50 |
熊岚雨 （吉林大学） |
Robust quasi-likelihood estimation for the negative binomial integer-valued GARCH(1,1) model with an application to transaction counts |
9:50-10:10 |
Tea Break |
主持人：梁汉营（同济大学） |
10:10-10:40 |
薛留根 （北京工业大学） |
Empirical likelihood in semiparametric models |
10:40-11:10 |
郑晨 （河南大学） |
基于统计的遥感大数据智能解译 |
11:10-11:40 |
胡玉萍 （郑州大学） |
响应变量随机缺失下部分函数型线性回归模型的统计推断 |
11:40 |
自助午餐，中州颐和酒店 |
2019年7月22日，星期一下午，数学院一楼报告厅 |
主持人：卢一强（信息工程大学） |
14:30-15:00 |
张军舰 （广西师范大学） |
均值变点的两阶段估计 |
15:00-15:30 |
陈华萍 （吉林大学） |
Binomial AR(1) processes with innovational outliers |
15:30-16:00 |
李哲源 （河南大学） |
分布型广义加性模型（GAMLSS）的基本理论和案例展示 |
16:00-16:20 |
Tea Break |
主持人：张军舰（广西师范大学） |
16:20-16:50 |
李本崇 （西安电子科技大学） |
Support condition for equivalent characterization of graph laws |
16:50-17:20 |
赵占平 （黄淮学院） |
Bayesian sample size determination for two Independent binomial experiments |
17:20-17:50 |
冯三营 （郑州大学） |
Testing for heteroskedasticity in two-way fixed effects panel data models |
17:50-18:20 |
解俊山 （河南大学） |
Asymptotic distribution of the maximum interpoint distance for high-dimensional data. |
18:30 |
晚宴，中州颐和酒店二楼 |
2019年7月23日，星期二上午，数学院二楼会议室 |
8:00-10:00 |
mk体育明伦校区参观考察 |
10:30-12:00 |
mk体育统计学科发展和博士、博士后培养座谈会 |
12:00 |
自助午餐，中州颐和酒店 |
|
|
|
|
Title and Abstracts
Title：Quantile regression for partially linear varying-coefficient model with censoring indicators missing at random
Speaker：梁汉营 (同济大学)
Abstract: In this talk, we focus on the partially linear varying-coefficient quantile regression model when the data are right censored and the censoring indicator is missing at random. Based on the calibration and imputation methods, a three-stage approach is proposed to construct the estimators of the linear part and the nonparametric varying-coefficient function for this model. At the same time, we discuss the variable selection of the covariates in the linear part by adopting adaptive LASSO penalty. Under appropriate assumptions, the asymptotic normality of the proposed estimators is established, and the penalized estimators are proven to have the oracle property. Simulation study and a real data analysis are conducted to evaluate the performance of the proposed estimators.
Title：Asymptotic efficiency of imputation based semi-supervised learning in misspecified setting
Speaker：吕绍高 (南京审计大学)
Abstract: We are concerned with statistical inferences on semi-supervised learning based on kernel methods in misspecified setting. Though semi-supervised learning has been extensively studied and many learning algorithms for semi-supervised problems have been proposed in the literature, there has been little theoretical justification under general setting on when and how the unlabeled data can be exploited to improve inference of a learning task. The goal of this paper is attempted to narrow the gap from the viewpoint of asymptotic variance of our proposed two-step estimation. Under the misspecified setting, the proposed pointwise nonparametric estimator has a smaller asymptotic variance than the supervised estimator using the labeled data alone.
Title：Robust quasi-likelihood estimation for the negative binomial integer-valued GARCH(1,1) model with an application to transaction counts
Speaker：熊岚雨 (吉林大学)
Abstract: For count time series analysis, the Poisson integer-valued generalized autoregressive conditional heteroscedastic model is very popular but is not usually suitable in the existence of potential extreme observations. Maximum likelihood estimator is commonly used to estimate parameters, but it is highly affected by the outliers. This paper has three main aims. First, we apply the negative binomial model in our study for count time series analysis and consider the maximum likelihood estimation of this model. Second, we extend the Mallows’ quasi-likelihood method proposed in the generalized linear models to our situation. Besides, we establish the consistency and asymptotic normality for the resulting robust estimators under some regularity conditions. Third, the performances of these robust estimators in the presence of transient shifts and additive outliers are investigated via simulations. We apply the robust estimator to two stock-market data sets and their prediction performances are assessed by in-sample and out-of-sample predictions.
Title：Empirical likelihood in semiparametric models
Speaker：薛留根 (北京工业大学)
Abstract: In this talk, we discuss the empirical likelihood based inference problem in semiparametric models. Firstly, we investigate the empirical likelihood based inference for the parameters in a partially linear single-index model. we propose a bias correction method to achieve that the empirical likelihood ratio has standard chi-square limit. Secondly, we investigate the empirical likelihood-based inference for a varying coefficient model with longitudinal data. we propose three empirical likelihood ratios: the naive empirical likelihood ratio, the mean-corrected empirical likelihood ratio and the residual-adjusted empirical likelihood ratio, and show that these ratios have chi-square limits. In addition, when some components are of particular interest, we suggest the mean-corrected and residual-adjusted partial empirical likelihood ratios for the construction of the confidence regions/intervals. A simulation study is undertaken to compare the empirical likelihood and the normal approximation methods in terms of coverage accuracies and average areas/widths of confidence regions/intervals. An example in epidemiology is used for illustration.
Title：基于统计的遥感大数据智能解译
Speaker：郑晨 (mk体育)
Abstract: 随着传感器与空间数据获取能力的进步，人类可得到的遥感数据迎来了爆炸式的增长。在海量的遥感影像数据中，探索和解决智能化的数据分析和信息提取，实现数据-信息-知识的自动转化已成为一个重要研究问题。由于遥感影像数据属于高随机信号，地物对象的特征和不同地物对象间的空间关系都具有较强的随机性，因此，基于统计的遥感数据分析方法受到了广泛的关注和研究。其中，马尔科夫随机场模型因其完备的理论基础及有效的空间描述能力受到了广泛关注，本报告将介绍马氏场模型的研究历史、发展现状及作者的相关工作。
Title：响应变量随机缺失下部分函数型线性回归模型的统计推断
Speaker：胡玉萍 (郑州大学)
Abstract: 本文主要目的是在响应变量 Y 随机缺失时, 构造部分函数型线性回归模型中未知参数的置信域, 并估计未知斜率函数. 提出了两种经验似然方法来完成研究目标.所提方法有其优点, 不仅克服了非参数模型选择最优带宽困难, 并且选择概率函数的估计是相和的. 进一步给出, 边际倾向得分加权经验对数似然比统计量渐近于标准卡方分布, 相应的极大经验似然估计具有渐近正态性.
Title：均值变点的两阶段估计
Speaker：张军舰 (广西师范大学)
Abstract: 变点问题是统计学中比较热门的一个研究方向，广泛应用于金融、经济、地质等领域。论文在分析均值变点LASSO估计的基础上，构造了最大累计偏差和统计量，进一步提出了基于逐个变点估计思想的两阶段变点估计方法。论文给出了实现该方法的具体算法并对其计算复杂度进行分析，指出该估计方法具有一致性等性质。其后用蒙特卡洛模拟方法对其进行验证，并与LASSO估计和SaRa估计进行比较。模拟结果表明估计具有一致性，其估计收敛速度优于其他两种方法。
Title：Binomial AR(1) processes with innovational outliers
Speaker：陈华萍 (吉林大学)
Abstract: Binomial integer-valued AR processes have been well studied in the literature, but there is little progress in modelling bounded integer-valued time series with outliers. In this paper, we first review some basic properties of the binomial integer-valued AR(1) process and then we introduce binomial integer-valued AR(1) processes with two classes of innovational outliers. We focus on the joint conditional least squares and the joint conditional maximum likelihood estimates of models' parameters and the probability of occurrence of the outlier. Their large-sample properties are illustrated by simulation studies. Artificial and real data examples are used to demonstrate good performances of the proposed models.
Title： 分布型广义加性模型（GAMLSS）的基本理论和案例展示
Speaker：李哲源 (mk体育)
Abstract: GAMLSS回归模型是广义加性模型（GAM）的一种推广。经典的广义线性或加性模型（GLM/GAM）只能对数据均值进行回归，但GAMLSS可以进一步对数据的方差、偏低和峰度进行回归。这一建模框架已经包含了比如异方差模型和零膨胀模型等热点，并还在继续扩展。本报告将简要介绍GAMLSS的基本理论模块，并进行案例展示。
Title：Support condition for equivalent characterization of graph laws
Speaker：李本崇 (西安电子科技大学)
Abstract: Structurally Markov distributions over a set of graphs which is also called graph laws complete the fully Markov Bayesian structure of model selection. In this paper, we characterize conditional independence structures induced by graphs in terms of closure operation. Further, we resolve the open question on support condition for equivalent characterization of graph laws posed in Byrne and Dawid (2015).
Title：Bayesian sample size determination for two Independent binomial experiments
Speaker：赵占平 (黄淮学院)
Abstract: Sample size determination is commonly encountered in modern medical studies for independent binomial experiments. A new approach for calculating sample size is developed by comparing Bayesian and frequentist idea when a hypothesis test between two binomial proportions is conducted. Sample size is calculated according to Bayesian posterior decision function and power of the most powerful test under 0-1 loss function. Sample are investigated for two cases that two proportions are equal to some fixed value or a random value. A simulation study and a real example are used to illustrate the proposed methodologies.
Title：Testing for heteroskedasticity in two-way fixed effects panel data models
Speaker：冯三营 (郑州大学)
Abstract: In this paper, we propose a new method for testing heteroskedasticity in two-way fixed effects panel data models under two important scenarios where the cross-sectional dimension is large and the temporal dimension is either large or fixed. Specifically, we will develop test statistics for both cases under the conditional moment framework, and derive their asymptotic distributions under both the null and alternative hypotheses. The proposed tests are distribution free and can easily be implemented using the simple auxiliary regressions. Simulation studies and two real data analyses demonstrate that our proposed tests perform well in practice, and may have the potential for wide application in econometric models with panel data.
Title：Asymptotic distribution of the maximum interpoint distance for high-dimensional data.
Speaker：解俊山(mk体育)
Abstract: Let be a random sample coming from a dimensional population with independent components. Denote the maximum interpoint Euclidean distance by . When both the dimension and the sample size tend to infinity, it proves that under a suitable normalization asymptotically obeys Gumbel distribution. The proofs are mainly depend on the Stein-Chen Poisson approximation and the moderate deviation of the partial sum of independent random variables.
