Limiting distributions of functionals of Markov chains

Let \s{X_n, n ? 0\s} and \s{Y_n, n ? 0\s} be two stochastic processes such that Y_n depends on X_n in a stationary manner, i.e. P(Y_n ? A\vbX_n) does not depend on n. Sufficient conditions are derived for Y_n to have a limiting distribution. If X_n is a Markov chain with stationary transition probabilities and Y_n = f(X_n,..., X_n+k) then Y_n depends on X_n is a stationary way. Two situations are considered: (i) \s{X_n, n ? 0\s} has a limiting distribution (ii) \s{X_n, n ? 0\s} does not have a limiting distribution and exits every finite set with probability 1. Several examples are considered including that of a non-homogeneous Poisson process with periodic rate function where we obtain the limiting distribution of the interevent times.     

Multivariate distributions having Weibull properties

Random variables X₁ ,…, X_n are said to have a joint distribution with Weibull minimums after arbitrary scaling if min_i(a_iX_i) has a one dimensional Weibull distribution for arbitrary constants a_i > 0, i = 1,…, n. Some properties of this class are demonstrated, and some examples are given which show the existence of a number of distributions belonging to the class. One of the properties is found to be useful for computing component reliability importance. The class is seen to contain an absolutely continuous Weibull distribution which can be generated from independent uniform and gamma distributions.     

On the Stability of Eaton’s Characterization by the Properties of Linear Forms

We consider a population and a sample X ₁,X ₂,…,X _n of n independent observations drawn from this population. We assume that two suitably chosen linear statistics of X ₁,X ₂,…,X _n are given. The assumption that these statistics are identically distributed or have the same distribution as the monomial X ₁ can be used to characterize various populations. This is an object of the so-called characterization theorems. But if the assumptions of a characterization theorem are fulfilled only approximately, then can we state that the conclusion of this characterization is also fulfilled approximately? Theorems concerning problems of this type are called stability theorems. By Eaton’s theorem, if, under additional conditions, two linear statistics $(X_{1}+\cdots +X_{k_{1}})/k_{1}^{1/\alpha}We consider a population and a sample X ₁,X ₂,…,X _n of n independent observations drawn from this population. We assume that two suitably chosen linear statistics of X ₁,X ₂,…,X _n are given. The assumption that these statistics are identically distributed or have the same distribution as the monomial X ₁ can be used to characterize various populations. This is an object of the so-called characterization theorems. But if the assumptions of a characterization theorem are fulfilled only approximately, then can we state that the conclusion of this characterization is also fulfilled approximately? Theorems concerning problems of this type are called stability theorems. By Eaton’s theorem, if, under additional conditions, two linear statistics and have the same distribution as the monomial X ₁, then this monomial has a symmetric stable distribution of order α. The stability estimation in this theorem is investigated in the λ ₀-metric.      

Exponential sums and rectangular partitions

Let F denote a finite field with q=p^f elements, and let σ(A) equal the trace of the square matrix A. This paper evaluates exponential sums of the form S(E,X₁,…,X_n)e{?σ(CX₁?X_nE)}, where S(E,X₁,…,X_n) denotes a summation over all matrices E,X₁,…,X_n of appropriate sizes over F, and C is a fixed matrix. This evaluation is then applied to the problem of counting ranked solutions to matrix equations of the form U₁?U_αA+DV₁?V_β=B where A,B,D are fixed matrices over F.     

On separating systems whose elements are sets of at most k elements

A system A₁,…,A_m of subsets of X?{1,…,n} is called a separating system if for any two distinct elements of X there is a set A_i (1?i?m) that contains exactly one of the two elements. We investigate separating systems where each set A_i has at most k elements and we are looking for minimal separating systems, that means separating systems with the least number of subsets. We call this least number m(n,k). Katona has proved good bounds on m(n,k) but his proof is very complicated. We give a shorter and easier proof. In doing so we slightly improve the upper bound of Katona.     

Optimal recovery of linear functionals on sets of finite dimension

Suppose that X is a linear space and L ₁, …, L _n is a system of linearly independent functionals on P, where P ? X is a bounded set of dimension n + 1. Suppose that the linear functional L ₀ is defined in X. In this paper, we find an algorithm that recovers the functional L ₀ on the set P with the least error among all linear algorithms using the information L ₁ f, …, L _n f, f ∈ P.     

Convergence in distribution of quotients of order statistics

Let X₁, X₂,… be i.i.d. random variables with continuous distribution function F < 1. It is known that if 1 - F(x) varies regularly of order - p, the successive quotients of the order statistics in decreasing order of X₁,…,X_n are asymptotically independent, as n→∞, with distribution functions x^kp, k = 1, 2, …. A strong converse is proved, viz. convergence in distribution of this type of one of the quotients implies regular varation of 1 - F(x).     

On covering a product of sets with products of their subsets

Suppose that S₁,…,S_N are collections of subsets of X₁,…,X_N, respectively, such that n_i subsets belonging to S_i, and no fewer, cover X_i for all i. the main result of this paper is that to cover X₁ x…x X_N requires no fewer than σ^N_i=1 (n_i–1) + 1 and no more than Π^N_i=1n_i subsets of the form A₁ x…x A_N, where A_i∈S₁foralli. Moreo ver, these bounds cannot be improved. Identical bounds for the spanning number of a normal product of graphs are also obtained.     

Bounds for the uniform deviation of empirical measures

If X₁,…,X_n are independent identically distributed R^d-valued random vectors with probability measure μ and empirical probability measure μ_n, and if $\text{a}$ is a subset of the Borel sets on R^d, then we show that P{sup_{A∈$\text{a}$}|μ_n(A)?μ(A)|≥ε} ≤ cs($\text{a}$, n²)e^?2n∈2, where c is an explicitly given constant, and s($\text{a}$, n) is the maximum over all (x₁,…,x_n) ∈ R^dn of the number of different sets in {{x₁…,x_n}∩A|A ∈$\text{a}$}. The bound strengthens a result due to Vapnik and Chervonenkis.     

On the average number of records for sequences of not identically distributed random variables

The scheme of n series of independent random variables X ₁₁, X ₂₁, …, X _k1, X ₁₂, X ₂₂, …, X _k2, …, X _1n , X _2n , …, X _kn is considered. Each of these successive series X _1m , X _2m , …, X _km , m = 1, 2, …, n consists of k variables with continuous distribution functions F ₁, F ₂, …, F _k , which are the same for all series. Let N(nk) be the number of upper records of the given nk random variables, and EN(nk) be the corresponding expected value. For EN(nk) exact upper and lower estimates are obtained. Examples are given of the sets of distribution functions for which these estimates are attained.     

A characterization of the multivariate normal distribution

Let X_j (j = 1,…,n) be i.i.d. random variables, and let Y′ = (Y₁,…,Y_m) and X′ = (X₁,…,X_n) be independently distributed, and A = (a_jk) be an n × n random coefficient matrix with a_jk = a_jk(Y) for j, k = 1,…,n. Consider the equation U = AX, Kingman and Graybill [Ann. Math. Statist.41 (1970)] have shown U ～ N(O,I) if and only if X ～ N(O,I). provided that certain conditions defined in terms of the a_jk are satisfied. The task of this paper is to delete the identical assumption on X₁,…,X_n and then generalize the results to the vector case. Furthermore, the condition of independence on the random components within each vector is relaxed, and also the question raised by the above authors is answered.     

TVaR-based capital allocation for multivariate compound distributions with positive continuous claim amounts

In this paper, we consider a portfolio of n dependent risks X₁,…,X_n and we study the stochastic behavior of the aggregate claim amount S=X₁+?+X_n. Our objective is to determine the amount of economic capital needed for the whole portfolio and to compute the amount of capital to be allocated to each risk X₁,…,X_n. To do so, we use a top-down approach. For (X₁,…,X_n), we consider risk models based on multivariate compound distributions defined with a multivariate counting distribution. We use the TVaR to evaluate the total capital requirement of the portfolio based on the distribution of S, and we use the TVaR-based capital allocation method to quantify the contribution of each risk. To simplify the presentation, the claim amounts are assumed to be continuously distributed. For multivariate compound distributions with continuous claim amounts, we provide general formulas for the cumulative distribution function of S, for the TVaR of S and the contribution to each risk. We obtain closed-form expressions for those quantities for multivariate compound distributions with gamma and mixed Erlang claim amounts. Finally, we treat in detail the multivariate compound Poisson distribution case. Numerical examples are provided in order to examine the impact of the dependence relation on the TVaR of S, the contribution to each risk of the portfolio, and the benefit of the aggregation of several risks.     

Asymptotics for the partial sum and its maximum of dependent random variables<Superscript>*</Superscript>

Let X ₁,…,X _n be pairwise asymptotically independent or pairwise upper extended negatively dependent real-valued random variables. Under the condition that the distribution of the maximum of X ₁,…,X _n belongs to some subclass of heavy-tailed distributions, we investigate the asymptotic behavior of the partial sum and its maximum generated by dependent X ₁,…,X _n . As an application, we consider a discrete-time risk model with insurance and financial risks and derive the asymptotics for the finite-time ruin probability.     

Strong law for the bootstrap

Let X₁, X₂, X₃, … be i.i.d. r.v. with E|X₁| < ∞, E X₁ = μ. Given a realization X = (X₁,X₂,…) and integers n and m, construct Y_n,i, i = 1, 2, …, m as i.i.d. r.v. with conditional distribution $\text{P}{}^1\text{(Y}{}_{\mathrm{n,i}}\text{= X}{}_{\mathrm{j}}\text{) =}\text{1}\text{n}$ for 1 ? j ? n. ($\text{P}{}^1$ denotes conditional distribution given X). Conditions relating the growth rate of m with n and the moments of X₁ are given to ensure the almost sure convergence of $\text{(}\text{1}\text{m}\text{)Σ}{}^{\mathrm{m}}{}_{\mathrm{i=1}}\text{Y}{}_{\mathrm{n,i}}$ toμ. This equation is of some relevance in the theory of Bootstrap as developed by Efron (1979) and Bickel and Freedman (1981).     

A pairwise independent stationary stochastic process

The purpose of this paper is to study pairwise independence in the context of strictly stationary stochastic processes {X_π, n = 0, ±1, …}. Our main result is an example of such a process that maximizes E(X₁X₂X₃). We also show that subject to some additional independence assumptions any two of these processes are distributionally the same. The spectral properties of this process are then analysed.     

Characterization of a class of bivariate distribution functions

Let F_X,Y(x,y) be a bivariate distribution function and P_n(x), Q_m(y), n, m = 0, 1, 2,…, the orthonormal polynomials of the two marginal distributions F_X(x) and F_Y(y), respectively. Some necessary conditions are derived for the co-efficients c_n, n = 0, 1, 2,…, if the conditional expectation E[P_n(X) ∥ Y] = c_nQ_n(Y) holds for n = 0, 1, 2,…. Several examples are given to show the application of these necessary conditions.     

Multilinear transformation on matrices

For each of several invariants P defined on M_n, the vector space of n-square matrices over a field, we determine the set of m-linear transformations $\text{φ:}\text{×}\text{1}\text{m}\text{M}{}_{\mathrm{n}}\text{→M}{}_{\mathrm{n}}$ which satisfy P(φ(X₁,…,X_m))=P(X₁?X_m) for all X₁,…,X_m∈M_n. Example: every multilinear determinant preserver is a product of linear determinant preservers.     

A generalized coupon collecting model as a parsimonious optimal stochastic assignment model

There is a given set of n boxes, numbered 1 thru n. Coupons are collected one at a time. Each coupon has a binary vector x ₁,…,x _n attached to it, with the interpretation being that the coupon is eligible to be put in box i if x _i =1,i=1…,n. After a coupon is collected, it is put in a box for which it is eligible. Assuming the successive coupon vectors are independent and identically distributed from a specified joint distribution, the initial problem of interest is to decide where to put successive coupons so as to stochastically minimize N, the number of coupons needed until all boxes have at least one coupon. When the coupon vector X ₁,…,X _n is a vector of independent random variables, we show, if P(X _i =1) is nondecreasing in i, that the policy π that always puts an arriving coupon in the smallest numbered empty box for which it is eligible is optimal. Efficient simulation procedures for estimating P _π (N>r) and E _π [N] are presented; and analytic bounds are determined in the independent case. We also consider the problem where rearrangements are allowed.     

On some properties of the Cesàro limit of a stochastic matrix

Let$\text{?(x}{}_1\text{,…,x}{}_{\mathrm{p}}\text{)}$ be a polynomial in the variables x1,…,xp with nonnegative real coefficients which sum to one, let A1,…,Ap be stochastic matrices, and let $\text{?}\text{?}\text{(A}{}_1\text{,…,A}{}_{\mathrm{p}}\text{)}$ be the stochastic matrix which is obtained from ? by substituting the Kronecker product of An11,…,Anppfor each term Xn11·?·Xnpp. In this paper, we present necessary and sufficient conditions for the Cesàro limit of the sequence of the powers of $\text{?}\text{?}\text{(A}{}_1\text{,…,A}{}_{\mathrm{p}}\text{)}$ to be equal to the Kronecker product of the Cesàro limits associated with each of A1,…,Ap. These conditions show that the equality of these two matrices depends only on the number of ergodic sets under$\text{?}\text{?}\text{(A}{}_1\text{,…,A}{}_{\mathrm{p}}\text{)}$ and?or the cyclic structure of the ergodic sets under A1,…,Ap, respectively. As a special case of these results, we obtain necessary and sufficient conditions for the interchangeability of the Kronecker product and the Cesàro limit operator.     

Absolute regularity and functions of Markov chains

We first give an extension of a theorem of Volkonskii and Rozanov characterizing the strictly stationary random sequences satisfying ‘absolute regularity’. Then a strictly stationary sequence {X_k, k = …, ?1, 0, 1,…} is constructed which is a 0?1 instantaneous function of an aperiodic Markov chain with countable irreducible state space, such that n^?2 var (X₁ + ? + X_n) approaches 0 arbitrarily slowly as n → ∞ and (X₁ + ? + X_n) is partially attracted to every infinitely divisible law.