Tribute to Endre Csáki and Pál Révész

Summary This article provides a glimpse of some of the highlights of the joint work of Endre Csáki and Pál Révész since 1979. The topics of this short exploration of the rich stochastic milieu of this inspiring collaboration revolve around Brownian motion, random walks and their long excursions, local times and additive functionals, iterated processes, almost sure local and global central limit theorems, integral functionals of geometric stochastic processes, favourite sites--favourite values and jump sizes for random walk and Brownian motion, random walking in a random scenery, and large void zones and occupation times for coalescing random walks.     

On the spectral norm of algebraic numbers

In this paper we continue to study the spectral norms and their completions ([4]) in the case of the algebraic closure $ \overline {\mathbb Q} $ of ? in ?. Let $ \widetilde{\overline{\mathbb{Q}}} $ be the completion of $ \overline {\mathbb Q} $ relative to the spectral norm. We prove that $ \widetilde{\overline{\mathbb{Q}}} $ can be identified with the R‐subalgebra of all symmetric functions of C(G), where C(G) denotes the ?‐Banach algebra of all continuous functions defined on the absolute Galois group G = Gal$ {\overline {\mathbb Q}} / {\mathbb Q} $. We prove that any compact, closed to conjugation subset of ? is the pseudo‐orbit of a suitable element of $ \widetilde{\overline{\mathbb{Q}}} $. We also prove that the topological closure of any algebraic number field in $ \widetilde{\overline{\mathbb{Q}}} $ is of the form $\widetilde{\mathbb{Q}[x]}$ with x in $ \widetilde{\overline{\mathbb{Q}}} $.     

ASYMPTOTIC NORMALITY OF KERNEL ESTIMATES OF A DENSITY FUNCTION UNDER ASSOCIATION DEPENDENCE

Let {X_n, n≥1} be a strictly stationary sequence of random variables, whichare either associated or negatively associated, f(·) be their common density. In this paper,the author shows a central limit theorem for a kernel estimate of f(·) under certain regularconditions.     

一种改进的进化规划算法

本对于全局优化问题提出一个改进的进化规划算法，该算法以概率p接收基于电磁理论求出合力方向作为随机搜索方向，以概率1-p接收按正态分布产生的随机搜索方向。改进算法不仅克服了传统进化规划算法随机搜索的盲目性，而且保留了传统进化规划算法全局搜索性。本算法应用于几个典型例题，数值结果表明本算法是可行的，有效的。     

期货经纪公司保证金的一种确定方法

本文基于马尔可夫链在存储论中的应用，结合Ergodic定理，得到确定期货经纪公司保证金的Ergodic模型，即一个双目标规划问题，然后应用乘积最大化准则，将该模型转化为单目标规划问题来求解。该方法考虑期货经纪公司承担的风险和对投资者的吸引程度，为保证金的确定提供新的思路。     

基于MMD算法的HMM参数重估计的矩阵表示

本文提出的MMD算法用于提高模型区别错误信息和正确信息的能力．利用该算法在对模型的参数进行重估计时．涉及到复杂的目标函数的梯度运算．击运用矩阵运算使得梯度运算变得简单明了，因此本文给出了MMD算法下的HMM参数重估计的矩阵表示形式并给出了证明．     

NCD系统的数学理论

无索赔折扣系统(No Claim Discount system,简记为NCD系统)是世界各国机动车辆险中广泛采用的一种经验费率厘定机制.本文尝试建立了NCD系统严谨的数学理论, 重点讨论了NCD系统的数学建模和稳态分析.此外,作为本文必要的数学前提,首先在第2节着重探讨了随机矩阵间的随机优序关系,并将所得结论运用至齐次不可约且遍历的马尔科夫链的研究中,这些内容也有其独立的数学上的兴趣.     

Inhomogeneous Voter Models in One Dimension

We establish conditions for survival and extinction of types of one-dimensional voter models, and show that increasing the flip rates at a finite number of sites typically does not affect survival, unless the flipping mechanism is altered. We provide an example of a modified voter model that does not survive but can be made to survive simply by altering the flip mechanism at one site. We also show that a rather general class of such models have clustering behavior.     

起源于引力场的Vaidya-Bonner-de Sitter黑洞的量子熵

利用brick-wall模型研究了引力场对Vaidya-Bonner-de Sitter黑洞熵的量子修正.当黑洞事 件视界不随超前时间变化时，结果与Reissner-Nordstrm-de Sitter黑洞的量子熵完全相 同.     

强非线性吸收下高斯光束Z-扫描衍射理论模型

 摘要：在同时考虑非线性折射和非线性吸收的情况下，借助光的衍射理论，研究了Z-扫描技术的衍射理论模型，分析了非线性折射和非线性吸收同时存在时Z-扫描曲线的形状特征和变化规律。数值计算表明，非线性吸收相移和非线性折射相移的比值是影响接收小孔输出光功率的重要因素。要想得到好的光限幅效果，应选取非线性相移比值远小于1或远大于2的非线性材料。