伪单调数列与带误差的凸组合迭代的收敛性

一 伪单调数列 定义1 设非负数列{ε_n}具有如下性质:“满足 a_(n+1)≤a_n+ε_n,n=1,2,… (1)且有下界的任意数列{a_n}必收敛”,则称数列{ε_n}具有“性质M”。 定理1 非负数列{ε_n}具有性质M的充要条件是级数sum from n=1 to ∞(ε_n)收敛。证 必要性:设非负数列{ε_n}具有性质M,取数列{a_(1n)}为     

两类惯量惟一的对称符号模式

§ 1　 IntroductionA sign pattern(matrix) A is a matrix whose entries are from the set{ +,-,0 } .De-note the setofall n× n sign patterns by Qn.Associated with each A=(aij)∈ Qnis a class ofreal matrices,called the qualitative class of A,defined byQ(A) ={ B =(bij)∈ Mn(R) |sign(bij) =aijfor all i and j} .　　 For a symmetric sign pattern A∈ Qn,by G(A) we mean the undirected graph of A,with vertex set { 1 ,...,n} and (i,j) is an edge if and only if aij≠ 0 .A sign pattern A∈ Qnis a do…     

由φ-混合序列产生的长程相依过程的广义强逼近定理

设{X_k;k≥1}是由X_k=∑_(i=0)~βα_iε_(k-i)所定义的滑动平均过程,其中{ε_i;-∞i∞}是一同分布的φ-混合相依变量序列,{α_i;i≥0}为满足条件α_i~i~(-α)l(i)的实数序列,l(i)为一缓变函数.当1/2α1时,{X_k;k≥1}为一长程相依过程.在Eε_0~2可能为无穷的条件下,对长程相依过程{X_k;k≥1}的部分和建立了一个更为一般性的强逼近定理.     

-混合误差下回归函数加权核估计的强相合性

设{yi}是固定在点{xi}的观察值,适合模型yi=g(xi) εi.其中g(x)是0,1上的未知函数,{εi}是均值为0的随机误差序列.文献中,在{εi}为独立同分布的条件下,通过构造新的函数gn(x),对g(x)进行了估计.论文将{εi}推广至~ρ-混合误差序列的情形,通过附加适当的条件和精细的计算,获得了用gn(x)估计g(x)的同样结论.     

扰动压缩条件的点列收敛性问题

研究完备度量空间X中满足ρ(xn,xn+1)≤Lρ(xn-1,xn)+εn的点列{xn}收敛性问题,其中L∈(0,1)为常数,εn非负是无穷小量称为扰动.文中的主要结论是:点列{xn}的收敛性由扰动εn决定,即当幂级数sum from n=1 to ∞εnxn的收敛半径R>1/L时,点列{xn}收敛.特别地,当R>1时,点列收敛;而R=1时,{xn}敛散性不能确定.     

局部时的Cauchy主值的精确完全收敛性

设{X,Xn;n≥1}为i.i.d.的随机变量序列,其均值为0且EX2=1.令s={Sn}n＞0为一维随机游动,其中S0=0,Sn=n∑k=1 Xk,对n≥1.定义G(n)为随机游动局部时的Cauchy主值.本文得到了,若存在某δ1＞0,E|X|2r/(3p-4)+δ1＜∞成立,那么对4/3＜p＜2及r＞p,有limε→02(r-p)/2-p∞Σn=1nr-2/p{│G(n)│εn1/p}=2p/(r-p)πE│N│2(R-P)/2-P∞ΣK=O(-1)K(2/2K+1)2(R-P)/2-P+1.     

(~ρ)-混合误差下回归函数加权核估计的强相合性

设{yi}是固定在点{xi}的观察值,适合模型yi=g(xi) εi.其中g(x)是[0,1]上的未知函数,{εi}是均值为0的随机误差序列.文献中,在{εi}为独立同分布的条件下,通过构造新的函数gn(x),对g(x)进行了估计.论文将{εi}推广至(~ρ)-混合误差序列的情形,通过附加适当的条件和精细的计算,获得了用gn(x)估计g(x)的同样结论.     

一类最优停止问题的解

Let {Z_i} be i.i.d.,and {ε_i} be i.i.d.,Bernoulli variables independent of {Z_i}.Set To≡z and T_n=ε_n(T_(n-1)-Z_n)for n≥1.Using Wiener-Hopf type equations,we give a newapproch to find optimal stopping rules for stopping T_n-nC,Where C>O.Special casesare nsidered in detail,some of them are difficult to treat by Ferguson's method.     

关于学生数学分析理解思考能力的一次测试试题

1 λ =supA且λ A ,则以下正确的是 :(A)存在单调增加数列an∈A使得limn→∞an=λ ;　 (B)存在数列an=λ -1n ∈A使得limn→∞an=λ ;(C)存在单调减少数列an∈A使得limn→∞an=λ ;　 (D)存在数列an=λ+1n ∈A使得limn→∞an=λ。2 数列 {an}满足 :对任意正数N ,存在正数ε,当n N时有 |an-a| ε,那么(A)数列 {an}收敛 ;　　　　　　　　　 (B)数列 {an}恒为常数 ;(C)数列 {an}当n充分大时恒为常数 ;(D)数列 {an}有界。3 limn→∞1nn !=0的证明 :(1 )对任意正数ε,limn→∞1n !εn=0 ;(2 )所以 ,存在N ,当n N时 ,1n !εn 1 ;(3 …     

A note on asymptotic behavior for negative drift random walk with dependent heavy-tailed steps and its application to risk theory

In this article, the dependent steps of a negative drift random walk are modelled as a two-sided linear process Xn =-μ ∞∑j=-∞ψn-jεj, where {ε, εn; -∞＜ n ＜ ∞}is a sequence of independent, identically distributed random variables with zero mean, μ＞0 is a constant and the coefficients {ψi;-∞＜ i ＜∞} satisfy 0 ＜∞∑j=-∞|jψj| ＜∞. Under the conditions that the distribution function of |ε| has dominated variation and ε satisfies certain tail balance conditions, the asymptotic behavior of P{supn≥0(-nμ ∞∑j=-∞εjβnj) ＞ x}is discussed. Then the result is applied to ultimate ruin probability.     

Scenery reconstruction with branching random walk

We study the problem of scenery reconstruction in arbitrary dimension using observations registered in boxes of size k (for k fixed), seen along a branching random walk. We prove that, using a large enough k for almost all the realizations of the branching random walk, almost all sceneries can be reconstructed up to equivalence.     

An iterative approximation procedure for the distribution of the maximum of a random walk

Let I(F) be the distribution function (d.f.) of the maximum of a random walk whose i.i.d. increments have the common d.f. F and a negative mean. We derive a recursive sequence of embedded random walks whose underlying d.f.'s F_k converge to the d.f. of the first ladder variable and satisfy FF₁F₂ on [0,∞) and I(F)=I(F₁)=I(F₂)=. Using these random walks we obtain improved upper bounds for the difference of I(F) and the d.f. of the maximum of the random walk after finitely many steps.     

Individual behaviors of oriented walks

Given an infinite sequence t=(_k)_k of −1 and +1, we consider the oriented walk defined by S_n(t)=∑_k=1ⁿ₁₂…_k. The set of t's whose behaviors satisfy S_n(t)bn^τ is considered ( and 0<τ1 being fixed) and its Hausdorff dimension is calculated. A two-dimensional model is also studied. A three-dimensional model is described, but the problem remains open.     

The mean number of sites visited by a pinned random walk

This paper concerns the number Z _n of sites visited up to time n by a random walk S _n having zero mean and moving on the d-dimensional square lattice Z ^d . Asymptotic evaluation of the conditional expectation of Z _n given that S ₀ = 0 and S _n = x is carried out under 2 + δ moment conditions (0 ≤ δ ≤ 2) in the cases d = 2, 3. It gives an explicit form of the leading term and reasonable estimates of the remainder term (depending on δ) valid uniformly in each parabolic region of (x, n). In the case x = 0 the problem has been studied for the simple random walk and its analogue for Brownian motion; the estimates obtained here are finer than or comparable to those found in previous works. Supported in part by Monbukagakusho grand-in-aid no. 15540109.     

Extremum of a time-inhomogeneous branching random walk

Consider a time-inhomogeneous branching random walk, generated by the point process L_n which composed by two independent parts: ‘branching’offspring X_n with the mean \mathrm{1}+{B(\mathrm{1}+n)}^{-\beta } for \beta \in (\mathrm{0},\mathrm{1}) and ‘displacement’ {{\displaystyle \xi }}_n with a drift {A(\mathrm{1}+n)}^{-\mathrm{2}\alpha } for \alpha \in (\mathrm{0},\mathrm{1}/\mathrm{2}), where the ‘branching’ process is supercritical for B>0 but ‘asymptotically critical’ and the drift of the ‘displacement’ {{\displaystyle \xi }}_n is strictly positive or negative for \left|A\right|\rangle \mathrm{0} but ‘asymptotically’ goes to zero as time goes to infinity. We find that the limit behavior of the minimal (or maximal) position of the branching random walk is sensitive to the ‘asymptotical’ parameter \beta and \alpha.     

Explicit Bounds for the Return Probability of Simple Random Walks

We give an exact computation of the second order term in the asymptotic expansion of the return probability, P_2n^d(0,0), of a simple random walk on the d-dimensional cubic lattice. We also give an explicit bound on the remainder. In particular, we show that P_2n^d(0,0) < 2 (d/4n)^d/2 where n M=M(d) is explicitly given.     

Tubular recurrence

We introduce the directed-edge-reinforced random walk and prove that the process is equivalent to a random walk in random environment. Using Oseledec"s multiplicative ergodic theorem, we obtain recurrence and transience criteria for random walks in random environment on graphs with a certain linear structure and apply them to directed-edge-reinforced random walks. This revised version was published online in August 2006 with corrections to the Cover Date.     

Parity properties and terminal points for lattice walks with steps of equal length

A lattice walk with all steps having the same length d is called a d-walk. In this note we examine parity properties of closed d-walks in RN in relation to N and d. We also characterize real numbers d for which in R² every lattice point can be the terminal point in a d-walk starting from the origin.     

First Hitting Times for Some Random Walks on Finite Groups

We consider a random walk on a finite group G based on a generating set that is a union of conjugacy classes. Let the nonnegative integer valued random variable T denote the first time the walk arrives at the identity element of G, if the starting point of the walk is uniformly distributed on G. Under suitable hypotheses, we show that the distribution function F of T is almost exponential, and we give an error term.     

The Arcsine Law

Let N _n denote the number of positive sums in the first n trials in a random walk (S _i) and let L _n denote the first time we obtain the maximum in S ₀,..., S _n. Then the classical equivalence principle states that N _n and L _n have the same distribution and the classical arcsine law gives necessary and sufficient condition for (1/n) L _n or (1/n) N _n to converge in law to the arcsine distribution. The objective of this note is to provide a simple and elementary proof of the arcsine law for a general class of integer valued random variables (T _n) and to provide a simple an elementary proof of the equivalence principle for a general class of integer valued random vectors (N _n, L _n).