��˹ͼģ�͵Ļ����������Ľ���IPSP�㷨

The IPSP algorithm is an efficient algorithm for computing maximum likelihood estimation of Gaussian graphical models. It first divides clique marginals of graphical models into several groups, and then it adjusts clique marginals in each group locally. This paper uses the IIPS algorithm on junction tree to replace local adjustment on each group in the IPSP algorithm and propose a resulting algorithm called IPSP-JT to reduce the complexity of the IPSP algorithm. Moreover, we give a graph with minimum edges used by IIPS to adjust locally, and we prove its existence and uniqueness and construct a local junction tree. Numerical experiments show that the IPSP-JT algorithm runs faster than the IPSP algorithm for large Gaussian graphical models.     

������ر���ԭ���·��ձ��ѵľ��鱴Ҷ˹����

Variance related premium principle is one of the most important principles not only in practice applications but also in research field of actuarial science. In this paper, the Bayesian models are established under variance related premium principle. The Bayesian estimate and credibility estimate of risk premium are derived. Furthermore, some statistical properties of estimators are discussed. In the models with multitude contract data, the unbiased consistent estimates of the structure parameters are proposed. Finally, the empirical Bayes estimator are proved to be asymptotically optimal.     

�������ռ��ж�����������IJ���ʽ

本文给出了上期望空间中独立随机变量部分和的最大不等式、指数 不等式、Marcinkiewicz-Zygmund不等式. 并且应用指数不等式和Marcinkiewicz-Zygmund不等式 研究了随机变量部分和序列完备收敛的性质.     

������ɢ�����γ̵���ɢ�ƽ�

??In this paper we describe the excursions from a set explicitly for
recurrent Markov chain with discrete time. A new exit system is presented through using a
law conditioned by specifying the starting point and ending point of excursions. In a simple
case, we verify that our conditioned excursion law is a discrete approximation for that of
a diffusion.     

ƽ����ʧ�����¾��з��������ṹ���Ŷ�ģ��

??In classical credibility theory, the claim amounts of different
insurance policies in a portfolio are assumed to be independent and the premiums are derived
under squared-error loss function. Wen et al. (2012) studied the credibility models with a
dependence structure among the claim amounts of one insurance policy that is called time
changeable effects and obtained the credibility formula. In this paper, we generalized this
dependence structure called time changeable effects to the claim amounts of different
insurance policies in a portfolio. Credibility premiums are obtained for Buhlmann and
Buhlmann-Straub credibility models with dependence structure under balanced loss function.     

����ȱʧ���ݵ�B������ָ��ģ�͹���

??In this paper, we studied the inverse probability weighted least squares estimation of single-index model with response variable missing at random. Firstly, the B-spline technique is used to approximate the unknown single-index function, and then the objective function is established based on the inverse probability weighted least squares method. By the two-stage Newton iterative algorithm, the estimation of index parameters and the B-spline coefficients can be obtained. Finally, through many simulation examples and a real data application, it can be concluded that the method proposed in this paper performs very well for moderate sample     

�ӷ�����ģ���¾�������ɾʧ���ݵĻع����

Clustered interval-censored failure time data often arises in medical studies when study subjects come from the same cluster. Furthermore, the failure time may be related to the cluster size. A simple and common approach is to simplify interval-censored data due to the lack of proper inference procedures for direct analysis. For this reason, we proposed the within-cluster resampling-based method to consider the case II interval-censored data under the additive hazards model. With-cluster resampling is simple but computationally intensive. A major advantage of the proposed approach is that the estimator can be easily implemented when the cluster size is informative. Asymptotic properties and some simulation results are provided and indicate that the proposed approach works well.     

�����ڷ����Ǹ����ݵ�ָ����������̬�ֲ�

本文研究了一种可用于分析非负数据的新的分布族. 它用一个随机表示来定义, 是一个半正态随机变量与一个指数随机变量的幂的混合. 其密度函数和性质, 包括风险函数, 矩和矩母函数被导出. 该分布族的实用性和灵活性通过一个实例用最大似然程序被阐明.     

ͬ�����������������ЧӦ�µij���ծȯ�����о�

本文在传统Lee-Carter人口死亡率模型的框架下, 引入同出生年人群死亡率之间的相关性效应, 从而对未来死亡率的动态变化进行更加具体的刻画. 同时借鉴Lin和Cox(2005)所提出的长寿债券构造机制, 基于中国的实际人口死亡率数据, 运用多维概率扭转变换对不完全市场下长寿债券的定价结果进行比较分析.     

����״̬�����������Եij���֤��

In this paper, we introduce the definitions of geometric strongly ergodic, strongly ergodic and weakly ergodic for continuous-state Markov chains, then we give a primary proof of equivalence of the ergodicities for continuous-state Markov chains.     

Complete Convergence for Weighted Sums of Negatively Superadditive Dependent Random Variables

Let $\{X_n,n\geq1\}$ be a sequence of negatively superadditive dependent (NSD, in short) random variables and $\{a_{nk}, 1\leq k\leq n, n\geq1\}$ be an array of real numbers. Under some suitable conditions, we present some results on complete convergence for weighted sums $\sum_{k=1}^na_{nk}X_k$ of NSD random variables by using the Rosenthal type inequality. The results obtained in the paper generalize some corresponding ones for independent random variables and negatively associated random variables.     

广义复合二项风险模型的若干大偏差结果

This paper is a further investigation of large deviation for partial and random sums of random variables, where {Xn,n ≥ 1} is non-negative independent identically distributed random variables with a common heavy-tailed distribution function F on the real line R and finite mean μ∈ R. {N（n）,n ≥ 0} is a binomial process with a parameter p ∈ （0,1） and independent of {Xn,n ≥ 1}; {M（n）,n ≥ 0} is a Poisson process with intensity λ 〉 0, Sn = ΣNn i=1 Xi-cM（n）. Suppose F ∈ C, we futher extend and improve some large deviation results. These results can apply to certain problems in insurance and finance.     

功率和的强逼近

设$\{\xi_n, n\geq 1\}$是正的随机变量序列, $\ep \xi_1=\theta>0$, 设$S_n = \sum\limits_{i=1}^n \xi_i, Y_n=n\theta\log (S_n/(n\theta))$. 在该文中, 当$\{\xi_n\}$是独立同分布或强平稳$\varphi$ -混合的正随机变量序列时,作者给出功率和$\{Y_n\}$用Wiener过程的强逼近结果.     

Local Precise Large and Moderate Deviations for Sums of Independent Random Variables

Let {X, X_k : k ≥ 1} be a sequence of independent and identically distributed random variables with a common distribution F. In this paper, the authors establish some results on the local precise large and moderate deviation probabilities for partial sums S_n =sum from i=1 to n(X_i) in a unified form in which X may be a random variable of an arbitrary type,which state that under some suitable conditions, for some constants T 0, a and τ 1/2and for every fixed γ 0, the relation P(S_n- na ∈(x, x + T ]) ~nF((x + a, x + a + T ]) holds uniformly for all x ≥γn~τ as n→∞, that is, P(Sn- na ∈(x, x + T ]) lim sup- 1 = 0.n→+∞x≥γnτnF((x + a, x + a + T ])The authors also discuss the case where X has an infinite mean.     

���帺������β��������ͼ������ֵβ���ʵĽ�����

Let $X_1,X_2,\ldots,X_n$ be a sequence of extended negatively dependent random variables with distributions $F_1,F_2,\ldots,F_n$,respectively. Denote by $S_n=X_1+X_2+\cdots+X_n$. This paper establishes the asymptotic relationship for the quantities $\pr(S_n>x)$, $\pr(\max\{X_1,X_2, \ldots,X_n\}>x)$, $\pr(\max\{S_1,S_2$, $\ldots,S_n\}>x)$ and $\tsm_{k=1}^n\pr(X_k>x)$ in the three heavy-tailed cases. Based on this, this paper also investigates the asymptotics for the tail probability of the maximum of randomly weighted sums, and checks its accuracy via Monte Carlo simulations. Finally, as an application to the discrete-time risk model with insurance and financial risks, the asymptotic estimate for the finite-time ruin probability is derived.     

Asymptotic Estimates of Finite-Time Ruin Probabilities with Dependent Risks and CMC Simulations

We consider a discrete-time risk model with dependence structures, where the claim-sizes \{X_n\}_{n\geq1} follow a one-sided linear process with independent and identically distributed (i.i.d.) innovations $\{\varepsilon_n\}_{n\geq1}$, and the innovations and financial risks form a sequence of independent and identically distributed copies of a random pair $(\varepsilon,Y)$ with dependent components. When the product \varepsilon Y has a heavy-tailed distribution, we establish some asymptotic estimates of the ruin probabilities in this discrete-time risk model. Finally, we use a Crude Monte Carlo (CMC) simulation to verify our results.     

和的乘积的重对数律

设{X,X_n,n≥1}是独立的或φ -混合的或 ρ -混合的正的平稳随机变量序列，或$\{X,X_n,n≥1}$是正的随机变量序列使得{X_n-EX,n≥1\} 是平稳遍历的鞅差序列，记S_n=\sum\limitsⁿ_{j=1}X_j, n≥1 . 该文在条件EX=μ> 0 及0 Var(X)<∞下，证明了部分和的乘积$\prod\limits^n_{j=1}S_j/n!\mu^n$在合适的正则化因子下的某种重对数律.     

NA样本的递归密度估计

设$\{X_n,n\geq 1\}$是一个严平稳的负相协的随机变量序列, 其概率密度函数为$f(x)$.本文讨论了$f(x)$的递归核估计量的联合渐近正态性.     

Asymptotic tail probability of randomly weighted sums of dependent random variables with dominated variation

This paper investigates the asymptotic behavior of tail probability of randomly weighted sums of dependent and real-valued random variables with dominated variation, where the weights form another sequence of nonnegative random variables. The result we obtain extends the corresponding result of Wang and Tang[7].     

同分布 ρ 混合序列的最大值不等式及其应用

设X_n,n≥1是同分布的ρ混合序列, 记S_n=∑ⁿ_{i=1 Xi}. 该文讨论了$\max\limits_{1\leq i\leq n}\frac{|S_i|}{i}$ $(n\geq1)$的分布函数的上界. 作为应用,获得了随机变量$\sup\limits_{n\geq1}\frac{|S_n|}{n}$的1阶矩及$p(>1)$阶矩分别存在有限的充分必要条件,这是一个与独立同分布场合相一致的结果.