Results 1  10
of
14
The twoparameter PoissonDirichlet distribution derived from a stable subordinator.
, 1995
"... The twoparameter PoissonDirichlet distribution, denoted pd(ff; `), is a distribution on the set of decreasing positive sequences with sum 1. The usual PoissonDirichlet distribution with a single parameter `, introduced by Kingman, is pd(0; `). Known properties of pd(0; `), including the Markov ..."
Abstract

Cited by 366 (33 self)
 Add to MetaCart
The twoparameter PoissonDirichlet distribution, denoted pd(ff; `), is a distribution on the set of decreasing positive sequences with sum 1. The usual PoissonDirichlet distribution with a single parameter `, introduced by Kingman, is pd(0; `). Known properties of pd(0; `), including the Markov chain description due to VershikShmidtIgnatov, are generalized to the twoparameter case. The sizebiased random permutation of pd(ff; `) is a simple residual allocation model proposed by Engen in the context of species diversity, and rediscovered by Perman and the authors in the study of excursions of Brownian motion and Bessel processes. For 0 ! ff ! 1, pd(ff; 0) is the asymptotic distribution of ranked lengths of excursions of a Markov chain away from a state whose recurrence time distribution is in the domain of attraction of a stable law of index ff. Formulae in this case trace back to work of Darling, Lamperti and Wendel in the 1950's and 60's. The distribution of ranked lengths of e...
Limiting behaviour of sums and the term of maximum modulus
 Proc. Lond. Math. Soc
, 1984
"... Suppose that {Xn: n ^ 1} are independent and identically distributed random variables with common continuous distribution function F. Set Sn = Xt +... + Xn and Mn = V" * i, and let X[l) be the term of maximum modulus, i.e. the Xt among Xu...,Xn for which  Xt  is largest. The influence of the ..."
Abstract

Cited by 9 (6 self)
 Add to MetaCart
(Show Context)
Suppose that {Xn: n ^ 1} are independent and identically distributed random variables with common continuous distribution function F. Set Sn = Xt +... + Xn and Mn = V" * i, and let X[l) be the term of maximum modulus, i.e. the Xt among Xu...,Xn for which  Xt  is largest. The influence of the extreme terms on the sample sum is studied by examining the behaviour of Sn/X[l) and Sn/Mn. The main results centre about conditions for these quantities to converge to 1 in probability and almost surely. Related results deal with ratios of order statistics and ratios of record values of {Xn}. A novel feature of our approach is to study the behaviour of {SB} between successive record values of {\Xn\}. 1. Introduction and
The quantiletransform–empiricalprocess approach to limit theorems for sums of order statistics
 Sums, trimmed sums and extremes, 215–267, Progr. Probab
, 1991
"... ..."
(Show Context)
Hypothesis testing for validation and certification
 J. Complexity
"... We develop a hypothesis testing framework for the formulation of the problems of 1) the validation of a simulation model and 2) using modeling to certify the performance of a physical system1. These results are used to solve the extrapolative validation and certification problems, namely problems wh ..."
Abstract

Cited by 4 (1 self)
 Add to MetaCart
(Show Context)
We develop a hypothesis testing framework for the formulation of the problems of 1) the validation of a simulation model and 2) using modeling to certify the performance of a physical system1. These results are used to solve the extrapolative validation and certification problems, namely problems where the regime of interest is different than the regime for which we have experimental data. We use concentration of measure theory to develop the tests and analyze their errors. This work was stimulated by the work of Lucas, Owhadi, and Ortiz [1] where a rigorous method of validation and certification is described and tested. In Remark 2.5 we describe the connection between the two approaches. Moreover, as mentioned in that work these results have important implications in the Quantification of Margins and Uncertainties (QMU) framework. In particular, in Remark 2.6 we describe how it provides a rigorous interpretation of the notion of confidence and new notions of margins and uncertainties which allow this interpretation. Since certain concentration parameters used in the above tests may be unkown, we furthermore show, in the last half of the paper, how to derive equally powerful tests which estimate them from sample data, thus replacing the assumption of the values of the concentration parameters with weaker assumptions. 1
A Higher Order Expansion for the Joint Density of the Sum and the Maximum with Applications to the Estimation of Climatological Trends
, 2000
"... The higher order expansion for the joint density of the sum and maximum of an iid sequence answers two questions: the theoretical question of what is the rate of the asymptotic independence between these two which was established by Chow and Teugels (1978) and Anderson and Turkman (1991), and the ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
The higher order expansion for the joint density of the sum and maximum of an iid sequence answers two questions: the theoretical question of what is the rate of the asymptotic independence between these two which was established by Chow and Teugels (1978) and Anderson and Turkman (1991), and the practical question of how to describe and model the dependence when the asymptotic result is not yet realized. Developing such an expansion under the three different domains of attraction for the maximum and modeling the annual total and maximum precipitation across the contiguous US, using a combined generalized extreme value (GEV) version of the expansion, are the key elements of this thesis. The three
Limit theorems for sums of order statistics
 in Sixth Int. Summer School in Probability Theory and Mathematical Statistics
, 1988
"... ..."
(Show Context)
Extreme Value Theory for Random Exponentials
"... Abstract. We study the limit distribution of upper extreme values of i.i.d. exponential samples {etXi} N i=1 as t → ∞, N → ∞. Two cases are considered: (A) ess sup X = 0 and (B) ess sup X = ∞. We assume that the function h(x) = − log P{X> x} (case B) or h(x) = − log P{X> −1/x} (case A) is ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
(Show Context)
Abstract. We study the limit distribution of upper extreme values of i.i.d. exponential samples {etXi} N i=1 as t → ∞, N → ∞. Two cases are considered: (A) ess sup X = 0 and (B) ess sup X = ∞. We assume that the function h(x) = − log P{X> x} (case B) or h(x) = − log P{X> −1/x} (case A) is (normalized) regularly varying at ∞ with index 1 < ϱ < ∞ (case B) or 0 < ϱ < ∞ (case A). The growth scale of N is chosen in the form N ∼ eλH0(t) (0 < λ < ∞), where H0(t) is a certain asymptotic version of the function H(t): = log E[etX] (case B) or H(t) = − log E[etX] (case A). As shown earlier by Ben Arous et al. [5], there are critical points λ1 < λ2, below which the LLN and CLT, respectively, break down, whereas for 0 < λ < λ2 the limit laws for the sum SN (t) = etX1 + · · · + etXN prove to be stable, with characteristic exponent α = α(ϱ, λ) ∈ (0, 2). In this paper, we obtain the (joint) limit distribution of the upper order statistics of the exponential sample. In particular, M1,N = max{etXi} N i=1 has asymptotically the Fréchet distribution with parameter α. We also show that the empirical extremal measure converges (in fdd) to a Poisson random measure with intensity d(x−α). These results are complemented by explicit representations of the joint limit distribution of SN (t) and M1,N (t) (and in particular of their ratio) in terms of i.i.d. random variables with standard exponential distribution. 1.
LOCALIZATION VS. DELOCALIZATION OF RANDOM DISCRETE MEASURES*
"... Abstract. Sequences of discrete measures µ(n) with random atoms {µ (n) i, i =1, 2,...} such that i µ(n) i = 1are considered. The notions of (complete) asymptotic localization vs. delocalization of such measures in the weak (mean or probability) and strong (with probability 1) sense are proposed and ..."
Abstract
 Add to MetaCart
(Show Context)
Abstract. Sequences of discrete measures µ(n) with random atoms {µ (n) i, i =1, 2,...} such that i µ(n) i = 1are considered. The notions of (complete) asymptotic localization vs. delocalization of such measures in the weak (mean or probability) and strong (with probability 1) sense are proposed and analyzed, proceeding from the standpoint of the largest atoms ’ behavior as n →∞. In this framework, the class of measures with the atoms of the form µ (n) i = Xi/Sn (i =1,...,n) is studied, where X1,X2,... is a sequence of positive, independent, identically distributed random variables (with a common distribution function F) and Sn = X1 + ···+ Xn. If E[X1] < ∞, then the law of large numbers implies that µ(n) is strongly delocalized. The case where E[X1] = ∞ is studied under the standard assumption that F has a regularly varying upper tail (with exponent 0 ≦ α ≦ 1). It is shown that for α<1, weak localization occurs. In the critical point α = 1, the weak delocalization is established. For α = 0, localization is strong unless the tail decay is “hardly slow.”
The Ratio of the Extreme to the Sum in a Random Sequence with Applications
, 1994
"... If X 1 , X 2 , . . . , X n is a sequence of nonnegative independent random variables with common distribution function F (t), we write X (n) for the maximum of the sequence and S n for its sum. The ratio variate R n = X (n) /S n is a quantity arising in the analysis of process speedup and the perf ..."
Abstract
 Add to MetaCart
If X 1 , X 2 , . . . , X n is a sequence of nonnegative independent random variables with common distribution function F (t), we write X (n) for the maximum of the sequence and S n for its sum. The ratio variate R n = X (n) /S n is a quantity arising in the analysis of process speedup and the performance of scheduling. O'Brien (1980) showed that R n 0 almost surely EX 1 < as n . Since {R n } is a uniformly bounded sequence it follows that EX 1 < ER n 0 as n . Here we show that, provided either that (i) EX 1 2 < or that (ii) 1  F(t) is a regularly varying function with index r < 1, it follows that ER n = ES n EX (n) ###### # # 1 + o(1) # # (n ) . Since the asymptotics of EX (n) is often readily calculated, this provides a useful estimate for the most significant behavior of the ratio R n in expectation. We apply this result to multiprocessor scheduling policies and to the behavior of sample statistics.
Limit Theorems on Ordered Random Vectors
"... Let X 1 ; X 2 ; ::: be independent identically distributed sdimensional random vectors, whose distribution belongs to the domain of attraction of a stable law. Let's X j;n ; j = 1; 2; ; n denote the order statistics built by increase of norms of random vectors X 1 ; X 2 ; ; X n ; i.e. j ..."
Abstract
 Add to MetaCart
Let X 1 ; X 2 ; ::: be independent identically distributed sdimensional random vectors, whose distribution belongs to the domain of attraction of a stable law. Let's X j;n ; j = 1; 2; ; n denote the order statistics built by increase of norms of random vectors X 1 ; X 2 ; ; X n ; i.e. jX 1;n j jX 2;n j jX n;n j: We investigate the asymptotic properties of random vectors T nk = (X 1n + +X n k;n )=jX n k+1;n j. 1 Introduction Let X 1 ; X 2 ; :::; X n ; ::: be independent identically distributed (i.i.d.) sdimensional random vectors having common absolute continuous distribution function. Let S(n) = X 1 +X 2 + ::: + X n ; F (x) = P (jX 1 j x); and p(x) be density of distribution of vector X 1 . sdimensional distribution is called to be stable if to every pair of vectors A 1 and A 2 and positive numbers B 1 and B 2 there always correspond a vector A and a positive number B such that for the three independent random vectors X, X 1 and X 2 possessing thi...