对称熵损失下成功概率p的E-Bayes估计

研究对称熵损失下成功概率p的Bayes估计和E-Bayes估计,证明了前者的存在性及唯一性.模拟结果表明E-Bayes估计优于极大似然估计和Bayes估计.并将E-Bayes方法应用在证券投资预测之中,预测效果较好.     

对称熵损失下成功概率P的E—Bayes估计

研究对称熵损失下成功概率P的Bayes估计和E—Bayes估计,证明了前者的存在性及唯一性．模拟结果表明E-Bayes估计优于极大似然估计和Bayes估计．并将E—Bayes方法应用在证券投资预测之中,预测效果较好．     

一个简化的新Laplace AR(1)模型 参数估计及其渐近性质

研究了一个简化的新的Laplace AR(1)模型参数的条件最小二乘估计和最大拟似然估计,并讨论了它们的强相合性和渐近正态性.通过数值模拟和实际例子,说明了最大拟似然估计及模型的优越性.     

Parameter Estimation for the NEAR（p） Model

As to the acronym NEAR(p), it means "New Exponential Autoregressive Process of order p". The NEAR(p) model is denned by where α0,α1,α2… are non-negative and sum to unity, and the residual sequence{εt} is defined as where q1, q2, … , qp+1 are non-negative and sum to unity, and {Et} is an independent and identically distributed (i.i.d.) sequence of standard exponential variants. Chan showed necessary and sufficient conditions for the existence of a stationary and ergodic NEAR(p) model.     

Median Unbiased Estimation of Bivariate Predictive Regression Models with Heavy-tailed or Heteroscedastic Errors

In this paper, we consider median unbiased estimation of bivariate predictive regression models with non-normal, heavy-tailed or heteroscedastic errors. We construct confidence intervals and median unbiased estimator for the parameter of interest. We show that the proposed estimator has better predictive potential than the usual least squares estimator via simulation. An empirical application to finance is given. And a possible extension of the estimation procedure to cointegration models is also described.     

基于删失加权广义第二类Beta分布的动态模型

若时间序列的观测值中存在大部分零值和一些正值,并且正值服从某一连续分布时,常见方法的拟合效果可能不太好.为此,2020年Harvey和Ito提出了一种删失方法,该方法将原分布向左移动,即将原随机变量减去一个常数,并将得到的负值赋值为0,但他们采用的广义第二类Beta分布有一定的局限性.文章考虑了更一般的加权广义第二类Beta分布,采用条件得分方法,提出了删失加权广义第二类Beta动态模型.文章将这个模型应用到澳大利亚日降雨量数据中,并与删失未加权、零增广加权、零增广未加权的广义第二类Beta动态模型相比较,发现删失加权的模型要优于其它三种模型.     

Softplus beta负二项整数值GARCH模型

INGARCH模型常基于泊松和负二项等分布来构造.Beta负二项(BNB)分布是一种灵活的分布,相关BNB-INGARCH模型最近被提出,该模型的条件均值是线性的,参数限制为非负的,不能建模负相关.本文首先提出对数线性BNB-INGARCH模型解决上述问题,但此模型不再具有线性均值的简单形式和类似ARMA的相关结构,采用softplus函数进一步构造了softplus BNB-INGARCH(p,q)模型作为主要研究对象.当p=q=1时证明了模型的平稳遍历性,给出了二阶矩存在的条件,并通过数值模拟验证该模型可以被线性近似,给出模型极大似然估计的相合性和渐近正态性,最后经过实际数据分析说明了模型的优良性.     

具有重尾或异方差误差的双变量预测回归模型的中位无偏估计

In this paper, we consider median unbiased estimation of bivariate predictive regression models with non-normal, heavy-tailed or heterescedastic errors. We construct confidence intervals and median unbiased estimator for the parameter of interest. We show that the proposed estimator has better predictive potential than the usual least squares estimator via simulation. An empirical application to finance is given. And a possible extension of the estimation procedure to cointegration models is also described.