Numerical algorithms for Panjer recursion by applying Bernstein approximation

In actuarial science, Panjer recursion (1981) is used in insurance to compute the loss distribution of the compound risk models. When the severity distribution is continuous with density function, numerical calculation for the compound distribution by applying Panjer recursion will involve an approximation of the integration. In order to simplify the numerical algorithms, we apply Bernstein approximation for the continuous severity distribution function and obtain approximated recursive equations, which are used for computing the approximated values of the compound distribution. The theoretical error bound for the approximation is also obtained. Numerical results show that our algorithm provides reliable results.     

Distributionally robust chance constrained problem under interval distribution information

This paper is concerned with distributionally robust chance constrained problem under interval distribution information. Using worst-case CVaR approximation, we present a tractable convex programming approximation for distributionally robust individual chance constrained problem under interval sets of mean and covariance information. We prove the worst-case CVaR approximation problem is an exact form of the distributionally robust individual chance constrained problem. Then, our result is applied to worst-case Value-at-Risk optimization problem. Moreover, we discuss the problem under several ambiguous distribution information and investigate tractable approximations for distributionally robust joint chance constrained problem. Finally, we provide an illustrative example to show our results.     

Approximation of the Ultimate Ruin Probability in the Classical Risk Model Using Erlang Mixtures

In this paper, we approximate the ultimate ruin probability in the Cramér-Lundberg risk model when claim sizes have an arbitrary continuous distribution. We propose two approximation methods, based on Erlang Mixtures, which can be used for claim sizes distribution both light and heavy tailed. Additionally, using a continuous version of the empirical distribution, we develop a third approximation which can be used when the claim sizes distribution is unknown and paves the way for a statistical application. Numerical examples for the gamma, Weibull and truncated Pareto distributions are provided.     

Sensitivity of risk measures with respect to the normal approximation of total claim distributions

A simple and commonly used method to approximate the total claim distribution of a (possibly weakly dependent) insurance collective is the normal approximation. In this article, we investigate the error made when the normal approximation is plugged in a fairly general distribution-invariant risk measure. We focus on the rate of convergence of the error relative to the number of clients, we specify the relative error’s asymptotic distribution, and we illustrate our results by means of a numerical example. Regarding the risk measure, we take into account distortion risk measures as well as distribution-invariant coherent risk measures.     

An approximation of F distribution by binomial probabilities

It is well known that the F distribution function can be expressed as an incomplete beta function. In particular when both degrees of freedom are even, the F distribution function may be simply expressed as a sum of binomial probabilities. In this note we consider the F distribution with at least one odd degree of freedom. By relating the F distribution to a generalized logistic distribution we obtain an approximation of the F distribution in terms of binomial probabilities. Because it is very close to the exact at the tails, the approximation is particularly suitable for estimating P-values of F ratio tests. All the computations involved can be carried out easily with only a hand calculator.     

On Approximating Lead Time Demand Distributions Using the Generalised λ-type Distribution

In this paper the use of the generalised λ-type distribution (GLD) is proposed for the analysis of standard inventory problems. Using this distribution to approximate the lead time demand distribution we analyse the generalised newsboy problem and a (Q, r) policy. The standard inventory measures like optimal order size, reorder level, average demand lost, etc. are obtained under the GLD and are compared with those given by Shore's approximation and also under exact distributional assumptions. Through a numerical study the various inventory measures are compared using the GLD and Shore's approximation with the exact distributions. The comparison reveals that the GLD approximation is better suited than Shore's approximation to model the lead time demand.     

A Saddlepoint Approximation to the Distribution of Inhomogeneous Discounted Compound Poisson Processes

In this article we propose an accurate approximation to the distribution of the discounted total claim amount, where the individual claim amounts are independent and identically distributed and the number of claims over a specified period is governed by an inhomogeneous Poisson process. More precisely, we compute cumulant generating functions of such discounted total claim amounts under various intensity functions and individual claim amount distributions, and invert them by the saddlepoint approximation. We provide precise conditions under which the saddlepoint approximation holds. The resulting approximation is numerically accurate, computationally fast and hence more efficient than Monte Carlo simulation.     

The not-quite non-atomic game: Normal approximation

Measure games with a large number of players are frequently approximated by nonatomic games. In fact, however, while it is true that values of large measure games will, under certain reasonable circumstances, converge to the value of a non-atomic game, it is also true that this convergence is quite slow. Using the multi-linear extension and the central limit theorem, we obtain an approximation which (because it is based on the normal distribution) we call the normal approximation. We show that, for two examples with several hundred and several thousand players respectively, the normal approximation is much better than the non-atomic approximation.     

Fluid models for many-server Markovian queues in a changing environment

Motivated by service systems with time-varying customer arrivals, we consider a fluid model as a macroscopic approximation for many-server Markovian queues alternating between underloaded and overloaded intervals. Our main result is a refinement of the piecewise stationary approximation (PSA) for the stationary distribution of the fluid model. The form of the refined approximation suggests simple metrics for assessing the accuracy of PSA for underloaded and overloaded intervals respectively.     

Skew-Gaussian random field

In this article, we introduce the concept of skewness to the Gaussian random field theory by defining a new two-dimensional non-Gaussian random field called skew-Gaussian random field. We derive the expected Euler characteristic of its excursion set. Moreover, an approximation to the size distribution of one connected component is derived. As an illustration, we present a simulation study to compare the approximation mechanism with the exact one.     

Asymptotic Approximations for the Distributions of the Multinomial Goodness-of-Fit Statistics under Local Alternatives

N. Cressie and T. R. C. Read (1984, J. Roy. Statist. Soc. B46, 440–464) introduced a class of multinomial goodness-of-fit statistics R^a based on power divergence. All R^a have the same chi-square limiting distribution under null hypothesis and have the same noncentral chi-square limiting distribution under local alternatives. In this paper, we investigate asymptotic approximations for the distributions of R^a under local alternatives. We obtain an expression of approximation for the distribution of R^a under local alternatives. The expression consists of continuous and discontinuous terms. Using the continuous term of the expression, we propose a new approximation of the power of R^a. We call the approximation AE approximation. By numerical investigation of the accuracy of the AE approximation, we present a range of sample size n that the omission of the discontinuous term exercises only slight influence on power approximation of R^a. We find that the AE approximation is effective for a much wider range of the value of a than the other power approximations, except for an approximation method which requires high computer performance.     

VALUE DISTRIBUTION THEORY AND DIOPHANTINE APPROXIMATION

In this paper, we will introduce some problems and results between Diophantine approximation and value distribution theory.     

A closed-form solution of the Black–Litterman model with conditional value at risk

We consider a portfolio optimization problem of the Black–Litterman type, in which we use the conditional value-at-risk (CVaR) as the risk measure and we use the multi-variate elliptical distributions, instead of the multi-variate normal distribution, to model the financial asset returns. We propose an approximation algorithm and establish the convergence results. Based on the approximation algorithm, we derive a closed-form solution of the portfolio optimization problems of the Black–Litterman type with CVaR.     

GBS Operators of Lupaş–Durrmeyer Type Based on Polya Distribution

In this paper, we study an extension of the bivariate Lupa?–Durrmeyer operators based on Polya distribution. For these operators we get a Voronovskaja type theorem and the order of approximation using Peetre’s K-functional. Then, we construct the Generalized Boolean Sum operators of Lupa?–Durrmeyer type and estimate the degree of approximation in terms of the mixed modulus of smoothness.     

On Stein?s method for infinite-dimensional Gaussian approximation in abstract Wiener spaces

In this paper, we generalize Stein?s method to “infinite-variate” normal approximation that is an infinite-dimensional approximation by abstract Wiener measures on a real separable Banach space. We first establish a Stein?s identity for abstract Wiener measures and solve the corresponding Stein?s equation. Then we will present a Gaussian approximation theorem using exchangeable pairs in an infinite-variate context. As an application, we will derive an explicit error bound of Gaussian approximation to the distribution of a sum of independent and identically distributed Banach space-valued random variables based on a Lindeberg-Lévy type limit theorem. In addition, an analogous of Berry-Esséen type estimate for abstract Wiener measures will be obtained.     

非参数回归函数最近邻估计误差分布的随机加权逼近

本文研究回归函数的最近邻估计的分布逼近问题．在一定条件下得到了最近邻回归估计误差的逼近分布,且逼近的精度比正态逼近精度更高．     

Distribution function estimates by Wasserstein metric and Bernstein approximation for C-1 functions

The aim of the paper is to estimate the density functions or distribution functions measured by Wasserstein metric, a typical kind of statistical distances, which is usually required in the statistical learningBased on the classical Bernstein approximation, a scheme is presented.To get the error estimates of the scheme, the problem turns to estimating the L1 norm of the Bernstein approximation for monotone C-1functions, which was rarely discussed in the classical approximation theoryFinally, we get a probability estimate by the statistical distance.     

Bayesian demand updating in the lost sales newsvendor problem: A two-moment approximation

We consider Bayesian updating of demand in a lost sales newsvendor model with censored observations. In a lost sales environment, where the arrival process is not recorded, the exact demand is not observed if it exceeds the beginning stock level, resulting in censored observations. Adopting a Bayesian approach for updating the demand distribution, we develop expressions for the exact posteriors starting with conjugate priors, for negative binomial, gamma, Poisson and normal distributions. Having shown that non-informative priors result in degenerate predictive densities except for negative binomial demand, we propose an approximation within the conjugate family by matching the first two moments of the posterior distribution. The conjugacy property of the priors also ensure analytical tractability and ease of computation in successive updates. In our numerical study, we show that the posteriors and the predictive demand distributions obtained exactly and with the approximation are very close to each other, and that the approximation works very well from both probabilistic and operational perspectives in a sequential updating setting as well.     

多组样本下GL－统计量的渐近性质

本文上要讨论多组样本下GL－统计量的渐近分布。这里我们使用了Gâteaut微分逼近方法,在多组i.i.d.样本下,给出了GL－统计量的渐近正态分布的一组条件,从而拓广了i.i.d.样本下GL－统计量的渐近正态分布的性质[1].     

最近邻回归估计的随机加权逼近

研究了一种最近邻回归估计的分布逼近问题，利用随机加权法，给出了最近邻回归估计误差的逼近分布及其逼近的精度，从而改进了文献「１」的结论。