On a new stopping rule for stochastic approximation

Summary A new stopping rule for the Robbins-Monro process, based on an F-statistic criterion is proposed and its asymptotic behavior established. On the basis of evidence obtained through experimental sampling, the procedure seems to work well over a wide variety of situations. A two-stage procedure, coupling the new rule with an earlier one proposed by Sielken [1973] is recommended for practical use.     

A stopping rule for the Robbins-Monro method

A stopping rule for the multidimensional Robbins-Monro stochastic approximation method is developed in this paper. Both moving average and stationary -mixing type of correlated noise processes are treated. Sequentially determined confidence ellipsoids are constructed to fulfill the goal for the determination of the stopping rule. The limit behavior of the algorithm is investigated. It is shown that the stopped Robbins-Monro process is asymptotically normal. Such asymptotic normality is established by means of weak convergence methods.Communicated by Y. C. Ho     

A stopping rule for symbolic dynamic filtering

One of the key issues in symbolic dynamic filtering (SDF) is how to obtain a lower bound on the length of symbol blocks for computing the state probability vectors of probabilistic finite-state automata (PFSA). Having specified an absolute error bound at a confidence level, this short work formulates a stopping rule by making use of Markov chain Monte Carlo (MCMC) computations.     

A stopping rule for choosing the best of three coins

A recently proposed stopping rule for choosing the best of three coins is presented as the first exit time from a certain polygonal region by a two-dimensional random walk. A diffusion approximation to the expected value of this exit time is developed, using techniques of conformal mapping. An algebraic method for calculating the exact value of this expectation is also described.     

Probabilistic characteristics of one stopping rule for steady control

A steady control procedure is considered. The expectation and the variance of the number of objects checked until the checking stops are found, if the checking procedure halts in the case where two defective objects appear among r last objects. Translated fromStatisticheskie Metody Otsenivaniya i Proverki Gipotez, pp. 80–83, Perm, 1990.     

A stopping rule for optimization using randomized search procedures

In many nonconvex programming problems, it is possible to locate local optima, but the global optimum may be difficult to determine. In such cases, a search procedure is often used, with random starting solutions, to find alternate local optima. This search can be terminated by a stopping rule, based upon Bayesian revised probability distributions, which determines the optimal number of iterations. The application of this rule to a resource allocation problem in project scheduling is illustrated.This work was supported in part by grants from the National Science Foundation and the Rochester Gas and Electric Corporation to the Massachusetts Institute of Technology.     

A random walk phenomenon under an interesting stopping rule

In the simple one-dimensional random walk setup, a path is described as follows. Toss a coin. If the result is head, score +1 and move one step forward; otherwise score ?1 and move one step backward. One is interested to know the position after a given number of steps. In this paper, once again a coin-tossing experiment is carried out. But this time, going by the philosophy of Bhanu et al. if score ?2, the result is head, otherwise score +1. The situation studied is the case that one decides to stop at the score of a given number n. Unlike their earlier case, the paper considers a general coin with probability of head equal to θ for 0?θ?et al. when θ?=?1/2.     

A stopping rule for a compound poisson random variable

An optimal empirical Bayesian stopping rule for the Poisson compounded with the geometric distribution is developed and applied to the problem of the sequential testing of computer software. For each checkpoint in time, either the software satisfies a desired economic criterion, or else the software testing is continued.     

A characterization of the uniform rule based on new robustness properties

We study the problem of allocating a divisible good among a group of people. Each person’s preferences are single-peaked. We consider situations in which there might be more of the resource to be assigned than was planned, or there might be less of the resource. Two robustness properties are formulated, which we call one-sided composition up and one-sided composition down. We show that only one rule satisfies irrelevance of null agents, the equal-division lower bound, and our robustness properties. This rule is the uniform rule.     

On a stopping rule for iterative orthogonalization of symmetric matrices

In this paper we consider three versions of Kovarik's iterative orthogonalization algorithms, for approximating the minimal norm solution of symmetric least squares problems. Although the convergence of these algorithms is linear, in practical applications we observed that a too big number of iterations can dramatically deteriorate the already obtained approximation. In this respect we analyse the above mentioned Kovarik–like methods according to the modifications they make on the “machine zero” eigenvalues of the problem (symmetric) matrix. We establish a theoretical almost optimal formula for the number of iterations necessary to obtain an enough accurate approximation, as well as to avoid the above mentioned troubles. Experiments on collocation discretization of a Fredholm first kind integral equation ilustrate the efficiency of our considerations. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)     

Some new properties of the Riesz probability distribution

The Riesz probability distribution on any symmetric cone and, in particular, on the cone of positive definite symmetric matrices represents an important generalization of the Wishart and of the matrix gamma distributions containing them as particular examples. The present paper is a continuation of the investigation of the properties of this probability distribution. We first establish a property of invariance of this probability distributions by a subgroup of the orthogonal group. We then show that the Pierce components of a Riesz random variable are independent, and we determine their probability distributions. Some moments and some useful expectations related to the Riesz probability distribution are also calculated. Copyright © 2017 John Wiley & Sons, Ltd.     

Householder矩阵的又一特性

给出了Householder矩阵的其它若干性质,利用本文中得到的正交向量组所对应的Householder矩阵的重要性质,解决了形如A=k1H1 k2H2 … knHn(ki∈R,Hi为n阶Householder矩阵,i=1,2,…n)的实对称阵的特性值与特征向量的问题,且任一实对称矩阵A均可表示为上述形式.     

Alternative optimality criteria of portfolio selection based upon threshold stopping rule

This paper proposes two types of alternative criteria of optimality for the continuous time portfolio selection problem. The optimality criteria, the so–called Laplace–Stieltjes transform (LST) criteria, are based on the assumption that the financial agent has a target level for the wealth accumulation process. These criteria are closely related to the so–called threshold stopping investment rule. We analytically derive the LST criteria and numerically compare them with the well–known Kelly criterion. It is shown that the portfolio strategies suggested may overcome the problem that the growth portfolio is often overestimated in several investment situations.     

Some new properties of biharmonic heat kernels

Contrary to the second-order case, biharmonic heat kernels are sign-changing. A deep knowledge of their behaviour may however allow us to prove positivity results for solutions of the Cauchy problem. We establish further properties of these kernels, we prove some Lorch–Szegö-type monotonicity results and we give some hints on how to obtain similar results for higher order polyharmonic parabolic problems.     

A note on optimal stopping of diffusions with a two-sided optimal rule

We consider a class of optimal stopping problems of diffusions with a two-sided optimal rule. We propose an approach for finding and characterizing the solution. We establish that the optimal stopping rule can be associated with the unique fixed point of an auxiliary function. The results are illustrated with an explicit example.     

Reliability properties under a stopping time shift

Considering a series representation of a coherent system using a shift transform of the components lifetime T_i, at its critical level Y_i, we study two problems. First, under such a shift transform, we analyse the preservation properties of the non-parametric distribution classes and secondly the association preserving property of the components lifetime under such transformations.