Estimator selection with respect to Hellinger-type risks

We observe a random measure N and aim at estimating its intensity s. This statistical framework allows to deal simultaneously with the problems of estimating a density, the marginals of a multivariate distribution, the mean of a random vector with nonnegative components and the intensity of a Poisson process. Our estimation strategy is based on estimator selection. Given a family of estimators of s based on the observation of N, we propose a selection rule, based on N as well, in view of selecting among these. Little assumption is made on the collection of estimators and their dependency with respect to the observation N need not be known. The procedure offers the possibility to deal with various problems among which model selection, convex aggregation and construction of T-estimators as studied recently in Birgé (Ann Inst H Poincaré Probab Stat 42(3):273?C325, 2006). For illustration, we shall consider the problems of estimation, complete variable selection and selection among linear estimators in possibly non-Gaussian regression settings.     

Order selection for heteroscedastic autoregression: A study on concentration

We consider an autoregressive model where the variance is allowed to be a function of time, unconditional on the past. Pötscher (1989) has proven that, regardless of the shape of the variance function, order selection can be made consistently. However, this procedure does not account for the non-stationary behavior. We consider the concentration of the variance function and its effect on order selection. We show that an order free estimate of the variance function can be constructed and propose an order selection criterion based on this estimate. Consistency is established and simulation results verify a large increase in the probability of selecting the correct order for finite samples.     

The distribution of the quasispecies for the Wright–Fisher model on the sharp peak landscape

We consider the classical Wright–Fisher model with mutation and selection. Mutations occur independently in each locus, and selection is performed according to the sharp peak landscape. In the asymptotic regime studied in Cerf (2014), a quasispecies is formed. We find explicitly the distribution of this quasispecies, which turns out to be the same distribution as for the Moran model.     

Portfolio selection based on upper and lower exponential possibility distributions

In this paper, two kinds of possibility distributions, namely, upper and lower possibility distributions are identified to reflect experts' knowledge in portfolio selection problems. Portfolio selection models based on these two kinds of distributions are formulated by quadratic programming problems. It can be said that a portfolio return based on the lower possibility distribution has smaller possibility spread than the one on the upper possibility distribution. In addition, a possibility risk can be defined as an interval given by the spreads of the portfolio returns from the upper and the lower possibility distributions to reflect the uncertainty in real investment problems. A numerical example of a portfolio selection problem is given to illustrate our proposed approaches.     

The effects of a new selection operator on the performance of a genetic algorithm

In this study, the new back controlled (BCSO) selection operator for genetic algorithm (GA) is presented. In the first stage, six existing operators and the BCSO were applied to the traveling salesman and space truss benchmark problems. An analysis of the results of the benchmark problems was made for each of the seven selection operators. In the second stage, the BCSO, the Roulette Wheel, dominant selection the hybrid II and operators were used to determine the fittest beam section has been used. In this stage, the cost of beam for a 18.3 m span developed by Mid-Atlantic States Pre-stressed Concrete Committee for Economic Fabrication and using a genetic algorithm the cost of a beam using BCSO were compared. It was determined that the beam produced using the proposed BCSO selection operator would be lower than those obtained with the existing selection operators. Also the determined beam cost is compares favourably with the average cost of a beam sold by the precast industry in the USA.     

Model selection method based on maximal information coefficient of residuals

The traditional model selection criterions try to make a balance between fitted error and model complexity. Assumptions on the distribution of the response or the noise, which may be misspecified, should be made before using the traditional ones. In this article, we give a new model selection criterion, based on the assumption that noise term in the model is independent with explanatory variables, of minimizing the association strength between regression residuals and the response, with fewer assumptions. Maximal Information Coefficient (MIC), a recently proposed dependence measure, captures a wide range of associations, and gives almost the same score to different type of relationships with equal noise, so MIC is used to measure the association strength. Furthermore, partial maximal information coefficient (PMIC) is introduced to capture the association between two variables removing a third controlling random variable. In addition, the definition of general partial relationship is given.     

MODEL SELECTION METHOD BASED ON MAXIMAL INFORMATION COEFFICIENT OF RESIDUALS

The traditional model selection criterions try to make a balance between fitted error and model complexity. Assumptions on the distribution of the response or the noise, which may be misspecified, should be made before using the traditional ones. In this ar- ticle, we give a new model selection criterion, based on the assumption that noise term in the model is independent with explanatory variables, of minimizing the association strength between regression residuals and the response, with fewer assumptions. Maximal Information Coe~cient （MIC）, a recently proposed dependence measure, captures a wide range of associ- ations, and gives almost the same score to different type of relationships with equal noise, so MIC is used to measure the association strength. Furthermore, partial maximal information coefficient （PMIC） is introduced to capture the association between two variables removing a third controlling random variable. In addition, the definition of general partial relationship is given.     

Contributions to multivariate analysis by Professor Yasunori Fujikoshi

The purpose of this article is to review the findings of Professor Fujikoshi which are primarily in multivariate analysis. He derived many asymptotic expansions for multivariate statistics which include MANOVA tests, dimensionality tests and latent roots under normality and nonnormality. He has made a large contribution in the study on theoretical accuracy for asymptotic expansions by deriving explicit error bounds. A large contribution has been also made in an important problem involving the selection of variables with introducing “no additional information hypotheses” in some multivariate models and the application of model selection criteria. Recently he is challenging to a high-dimensional statistical problem. He has been involved in other topics in multivariate analysis, such as power comparison of a class of tests, monotone transformations with improved approximations, etc.     

The full-information best choice problem with a random number of observations

The full-information best choice problem with a random number of observations is considered. N i.i.d. random variables with a known continuous distribution are observed sequentially with the object of selecting the largest. Neither recall nor uncertainty of selection is allowed and one choice must be made. In this paper the number N of observations is random with a known distribution. The structure of the stopping set is investigated. A class of distributions of N (which contains in particular the uniform, negative-binomial and Poisson distributions) is determined, for which the so-called “monotone case” occurs. The theoretical solution for the monotone case is considered. In the case where N is geometric the optimal solution is presented and the probability of winning worked out. Finally, the case where N is uniform is examined. A simple asymptotically optimal stopping rule is found and the asymptotic probability of winning is obtained.     

Order statistics and linear combination of order statistics arising from a bivariate selection normal distribution

In this paper, we derive distributions of order statistics and a linear combination of order statistics arising from a bivariate selection normal distribution. We show that they are mixtures of univariate selection normal distributions. We then illustrate the usefulness of these results with a real-life data relating to visual acuity analysis. Finally, some concluding remarks are made.     

基于条件分布函数的高维回归模型的变量选择方法

高维回归分析的变量选择问题是目前统计学研究的一个热点和难点问题.提出了一个基于条件分布函数的相关性度量准则,并在此基础上提出三种变量选择方法.与现有的方法相比,提出的方法不依赖于统计模型,可以适用于线性模型和非参数可加模型.数值模拟结果表明,即使协变量之间存在一定的相关性,方法也有较为满意的表现.     

An enhanced two-quantile Wilks methodology for engineering uncertainty analysis

Complex computational engineering uncertainty analyses have become more prevalent. When input parameters of such engineering models are uncertain, the output metric's uncertainty distribution is of an unknown parametric form. Since Wilks' method, named after the seminal paper by SS Wilks in 1941 entitled “Determination of sample sizes for setting tolerance limits”, is a nonparametric statistical procedure, it has received renewed interest, in particular in nuclear and chemical safety engineering. Unfortunately, the prevailing Wilks' method applied relies on arbitrary specification of order statistics' ranks with undue influence on the sample size recommendations that follow. Herein, a novel modification of Wilks' method involving two quantiles is proposed resolving that arbitrary rank selection. Together with a confidence level to be exceeded, these quantiles uniquely determine the parameters of an order statistics' beta distribution which drive the selection of symmetric tolerance limits. The modified procedure is demonstrated in two illustrative engineering uncertainty analysis examples drawn from the nuclear and chemical engineering domains.     

Testing for a multivariate generalized Pareto distribution

It has recently been shown by Rootzén and Tajvidi (Bernoulli, 12:917–930, 2006) that modelling exceedances of a random variable over a high threshold (peaks-over-threshold approach [POT]) can also in the multivariate setup be done rationally only by a multivariate generalized Pareto distribution (GPD). The selection of a proper threshold is, however, a crucial problem. The contribution of this paper is twofold: We develop first a non asymptotic and exact level-α test based on the single-sample t-test, which checks whether multivariate data are actually generated by a multivariate GPD. Secondly, this procedure is utilized for the derivation of a t-test based threshold selection rule in multivariate peaks-over-threshold models. The application to a hydrological data set illustrates this approach.      

Inhomogeneous maps and mathematical theory of selection

In this paper we develop a theory of general selection systems with discrete time and explore the evolution of selection systems, in particular, inhomogeneous populations. We show that the knowledge of the initial distribution of the selection system allows us to determine explicitly the system distribution at the entire time interval. All statistical characteristics of interest, such as mean values of the fitness or any trait can be predicted effectively for indefinite time and these predictions dramatically depend on the initial distribution. The Fisher Fundamental theorem of natural selection (FTNS) and more general the Price equations are the famous results of the mathematical selection theory. We show that the problem of dynamic insufficiency for the Price equations and for the FTNS can be resolved within the framework of selection systems. Effective formulas for solutions of the Price equations and for the FTNS are derived. Applications of the developed theory to some other problems of mathematical biology (dynamics of inhomogeneous logistic and Ricker model, selection in rotifer populations) are also given. Complex behavior of the total population size, the mean fitness (in contrast to the plain FTNS) and other traits is possible for inhomogeneous populations with density-dependent fitness. The temporary dynamics of these quantities can be investigated with the help of suggested methods.     

A Comparative Investigation on Model Selection in Independent Factor Analysis

With uncorrelated Gaussian factors extended to mutually independent factors beyond Gaussian, the conventional factor analysis is extended to what is recently called independent factor analysis. Typically, it is called binary factor analysis (BFA) when the factors are binary and called non-Gaussian factor analysis (NFA) when the factors are from real non-Gaussian distributions. A crucial issue in both BFA and NFA is the determination of the number of factors. In the literature of statistics, there are a number of model selection criteria that can be used for this purpose. Also, the Bayesian Ying-Yang (BYY) harmony learning provides a new principle for this purpose. This paper further investigates BYY harmony learning in comparison with existing typical criteria, including Akaik’s information criterion (AIC), the consistent Akaike’s information criterion (CAIC), the Bayesian inference criterion (BIC), and the cross-validation (CV) criterion on selection of the number of factors. This comparative study is made via experiments on the data sets with different sample sizes, data space dimensions, noise variances, and hidden factors numbers. Experiments have shown that for both BFA and NFA, in most cases BIC outperforms AIC, CAIC, and CV while the BYY criterion is either comparable with or better than BIC. In consideration of the fact that the selection by these criteria has to be implemented at the second stage based on a set of candidate models which have to be obtained at the first stage of parameter learning, while BYY harmony learning can provide not only a new class of criteria implemented in a similar way but also a new family of algorithms that perform parameter learning at the first stage with automated model selection, BYY harmony learning is more preferred since computing costs can be saved significantly.     

泛函型统计量的随机加权逼近

§ 1 引言及一般结果设X_,X_2,…X_n是来自分布为F(未知)的总体的n个iid.样本。它们的经验分布为 F_n(x)=1/n sum from i=1 to n I{X_ix}。 (1)常用的许多统计量都是通过经验分布而依赖于样本的,即能写成经验分布的泛函形式     

基于矩分析的资产组合选择

在风险资产收益分布为非正态的情景下,通过矩分析,研究其收益的高阶矩对资产组合选择的影响.首先,假设风险资产收益存在有限阶矩,泰勒展开边际财富期望效用,获得静态资产组合选择的近似解;其次,假设收益过程的跳跃产生收益分布的非正态性,运用随机控制方法获得动态资产组合选择的近似解析解,从高阶矩角度解释其特征。分析表明,超出峰度的存在导致减少风险资产投资,正(负)的偏度导致增加(减少)风险资产投资,该影响性随着它们及风险规避系数的增大而增强;可预测性导致资产组合存在正或负的对冲需求,取决于相关系数的符号和风险规避系数;跳跃性总体上减少风险资产投资;可预测性和跳跃性对动态资产组合选择的影响具有内在关联性。     

A simulation study for the distribution law of relative moments of evolution

Nine selection‐survival strategies were implemented in a genetic algorithm experiment, and differences in terms of evolution were assessed. The moments of evolution (expressed as generation numbers) were recorded in a contingency of three strategies (i.e., proportional, tournament, and deterministic) for two moments (i.e., selection for crossover and mutation and survival for replacement). The experiment was conducted for the first 20,000 generations in 46 independent runs. The relative moments of evolution (where evolution was defined as a significant increase in the determination coefficient relative to the previous generation) when any selection‐survival strategy was used fit a Log‐Pearson type III distribution. Moreover, when distributions were compared to one another, functional relationships were identified between the population parameters, revealing a degeneration of the Log‐Pearson type III distribution in a one‐parametrical distribution that can be assigned to the chosen variable—evolution strategy. The obtained theoretical population distribution allowed comparison of the selection‐survival strategies that were used. © 2012 Wiley Periodicals, Inc. Complexity, 2012.     

PROMETHEE-MD-2T method for project selection

Project selection is a real problem of multicriteria group decision making (MCGDM) where each decision maker expresses his/her preferences depending on the nature of the alternatives and on his/her own knowledge over them. Thus, information, as much quantitative as qualitative, coexists. The traditional methods of MCGDM developed for project selection usually discriminates in favour of quantitative information at the expense of qualitative information, and this is due to the capability to integrate this first type of information inside their procedure. In this article, two new multicriteria 2-tuple group decision methods called “Preference Ranking Organisation Method for Enrichment Evaluation Multi Decision maker 2-Tuple-I and II” (PROMETHEE-MD-2T-I and II) are presented. They are able to integrate inside their procedure both quantitative and qualitative information in an uncertain context. This has been performed by integrating a 2-tuple linguistic representation model dealing with non-homogeneous and imprecise information data made up by valued intervals, numerical and linguistic values into the aggregation operators of Promethee methods. Although they have been developed for project selection problems, these proposed methods can be applied to all kinds of decision-making problems with heterogeneous and multigranular information.     

FORECASTING TREE SIZE DISTRIBUTION BY NUMERICAL APPROXIMATION OF THE FORWARD KOLMOGOROV EQUATION

ABSTRACT. Forest management planning of uneven‐aged stands involves forecasting of the tree size distribution. The temporal development of the size distribution in a forest stand may be described by the forward Kolmogorov equation. The objective of this study is to illustrate that numerical approximation of the solution to the equation provides a reasonably accurate way of forecasting future tree size distribution, especially for stands with non‐normal size distribution. Furthermore, a method for the practical application is devised. The analyses compare observed and forecasted tree size distributions for two forest stands, 1) an unthinned stand of Sitka spruce (Picea sitchensis (Bong.) Carr.), and 2) an uneven‐aged stand of beech (Fagus sylvatica L.) managed under the selection system. The analyses show that the size distribution in the uneven‐aged stand may be forecasted correctly for a 20 25 year period, while for the even‐aged stand the method seems to fail after 10 to 15 years.