Bayesian Variable and Transformation Selection in Linear Regression

This article suggests a method for variable and transformation selection based on posterior probabilities. Our approach allows for consideration of all possible combinations of untransformed and transformed predictors along with transformed and untransformed versions of the response. To transform the predictors in the model, we use a change-point model, or “change-point transformation,” which can yield more interpretable models and transformations than the standard Box–Tidwell approach. We also address the problem of model uncertainty in the selection of models. By averaging over models, we account for the uncertainty inherent in inference based on a single model chosen from the set of models under consideration. We use a Markov chain Monte Carlo model composition (MC³) method which allows us to average over linear regression models when the space of models under consideration is very large. This considers the selection of variables and transformations at the same time. In an example, we show that model averaging improves predictive performance as compared with any single model that might reasonably be selected, both in terms of overall predictive score and of the coverage of prediction intervals. Software to apply the proposed methodology is available via StatLib.     

A review of credibilistic portfolio selection

This paper reviews the credibilistic portfolio selection approaches which deal with fuzzy portfolio selection problem based on credibility measure. The reason for choosing credibility measure is given. Several mathematical definitions of risk of an investment in the portfolio are introduced. Some credibilistic portfolio selection models are presented, including mean-risk model, mean-variance model, mean-semivariance model, credibility maximization model, α-return maximization model, entropy optimization model and game models. A hybrid intelligent algorithm for solving the optimization models is documented. In addition, as extensions of credibilistic portfolio selection approaches, the paper also gives a brief review of some hybrid portfolio selection models.     

Modeling and simulations of the clutch dynamics using approximations of the resulting friction forces

The work presents the effectiveness of a certain class of approximate models of the resulting friction force found during simulations of a simplified model of clutch dynamics. The friction models are based on integral expressions assuming fully developed sliding and Coulomb's classical law of friction on each element of the planar contact. Special approximations of the integral model of the friction force and the moment are proposed, which are based on Padé approximants and their generalizations. The system of clutch dynamics is simplified to a friction disk on a rotating master disk. Two different configurations are investigated, including the coaxial and non-coaxial arrangement of the disks. The model based on the generalization of Padé approximants is compared with the corresponding approximants based on the Taylor series expansion and with the model using the integral expressions for the resultant friction force and torque components.     

Genetic algorithms for portfolio selection problems with minimum transaction lots

Conventionally, portfolio selection problems are solved with quadratic or linear programming models. However, the solutions obtained by these methods are in real numbers and difficult to implement because each asset usually has its minimum transaction lot. Methods considering minimum transaction lots were developed based on some linear portfolio optimization models. However, no study has ever investigated the minimum transaction lot problem in portfolio optimization based on Markowitz’ model, which is probably the most well-known and widely used. Based on Markowitz’ model, this study presents three possible models for portfolio selection problems with minimum transaction lots, and devises corresponding genetic algorithms to obtain the solutions. The results of the empirical study show that the portfolios obtained using the proposed algorithms are very close to the efficient frontier, indicating that the proposed method can obtain near optimal and also practically feasible solutions to the portfolio selection problem in an acceptable short time. One model that is based on a fuzzy multi-objective decision-making approach is highly recommended because of its adaptability and simplicity.     

A power law type of time preference on intertemporal choices

Discount utility, based on utility theory, is used to study human decision behaviors under the consideration of time preference. It assumes that by means of the axiomatic system of rationality, it is possible to quantify human beings’ utilities by some explicable models for intertemporal decision making. Recent studies have been based on two basic models: the exponential and the hyperbolic discount models. These two types of model have been proved to be either too fast for discounting or too restricted for fitting human beings’ discounting behaviors. In this study, a power law discount model is proposed. Axiomatic approach is used to ascertain the existence of the power law discount utility, and empirical investigations are implemented to verify the effectiveness of the proposed model.     

Delta hedging strategies comparison

In this paper we implement dynamic delta hedging strategies based on several option pricing models. We analyze different subordinated option pricing models and we examine delta hedging costs using ex-post daily prices of S&P 500. Furthermore, we compare the performance of each subordinated model with the Black–Scholes model.     

Differential equation of a fractal electrode–electrolyte interface

Understanding an electrode–electrolyte interface (EEI) behavior is a valuable tool in several areas of science. There are models based on discrete fractal structures, which explain the measurements of linear and non-linear impedance at fixed frequencies, or at determined ranges of high and low current densities. A level by level discrete calculation is needed to evaluate these models, or the use of black-box models, which affect the good understanding of the phenomenon. A continuous model based on a differential equation of an EEI is presented in this paper. It includes an electrical circuit similar to a long transmission line. It has been deduced from the discrete Liu model.     

Robust tests based on dual divergence estimators and saddlepoint approximations

This paper is devoted to robust hypothesis testing based on saddlepoint approximations in the framework of general parametric models. As is known, two main problems can arise when using classical tests. First, the models are approximations of reality and slight deviations from them can lead to unreliable results when using classical tests based on these models. Then, even if a model is correctly chosen, the classical tests are based on first order asymptotic theory. This can lead to inaccurate p-values when the sample size is moderate or small. To overcome these problems, robust tests based on dual divergence estimators and saddlepoint approximations, with good performances in small samples, are proposed.     

Comparing discrete-event simulation and system dynamics: users' perceptions

This paper is the first of its type in that it provides an empirical study comparing the two simulation approaches of discrete-event simulation (DES) and system dynamics (SD). Prior comparison work is limited and mostly based on the authors' personal opinions. In the present work, the comparison is based on managers' (executive MBA students) perceptions of two simulation models of the same problem, one in DES and one in SD. The study found that there is no significant difference from the users' point of view between DES and SD in terms of model understanding and model usefulness. Some minor differences were found in terms of complexity and validity of the models, and the model results. The implications of our findings regarding model understanding, model complexity, model validity, model usefulness and model results are discussed.     

Justifying the small‐world phenomenon via random recursive trees

We present a new technique for proving logarithmic upper bounds for diameters of evolving random graph models, which is based on defining a coupling between random graphs and variants of random recursive trees. The advantage of the technique is three‐fold: it is quite simple and provides short proofs, it is applicable to a broad variety of models including those incorporating preferential attachment, and it provides bounds with small constants. We illustrate this by proving, for the first time, logarithmic upper bounds for the diameters of the following well known models: the forest fire model, the copying model, the PageRank‐based selection model, the Aiello‐Chung‐Lu models, the generalized linear preference model, directed scale‐free graphs, the Cooper‐Frieze model, and random unordered increasing k‐trees. Our results shed light on why the small‐world phenomenon is observed in so many real‐world graphs. © 2016 Wiley Periodicals, Inc. Random Struct. Alg., 50, 201–224, 2017     

Discrete grey forecasting model and its optimization

Although the grey forecasting model has been successfully adopted in various fields and demonstrated promising results, the literatures show its performance could be further improved. For this purpose, this paper proposes a novel discrete grey forecasting model termed DGM model and a series of optimized models of DGM. This paper modifies the algorithm of GM(1, 1) model to enhance the tendency catching ability. The relationship between the two models and the forecasting precision of DGM model based on the pure index sequence is discussed. And further studies on three basic forms and three optimized forms of DGM model are also discussed. As shown in the results, the proposed model and its optimized models can increase the prediction accuracy. When the system is stable approximately, DGM model and the optimized models can effectively predict the developing system. This work contributes significantly to improve grey forecasting theory and proposes more novel grey forecasting models.     

基于参数模型的EVaR风险度量计算方法及实证研究

针对EVaR(Expectile-based Value at Risk)风险度量提出了基于GARCH类和SV波动率模型的EVaR风险度量计算方法,即EVaR计算的参数模型方法.并基于模拟学生t分布时间序列数据,给出EVaR样本外预测的失败率检验方法:Kupiec失败率检验和动态分位数(DQ)检验法.与采用CARE(Conditional Autoregressive Expectile)模型的EVaR计算方法进行了对比研究,结果表明基于GARCH类模型和SV模型相对于基于CARE模型有更优的EVaR预测效果.选取2004年1月5日到2009年12月30日的国内外五个股票市场指数数据,针对日对数收益率进行了EVaR风险度量的实证研究,得出在金融危机期间,基于参数模型的EVaR预测要比基于CARE模型的EVaR预测更接近市场实际风险.     

Modeling international investment decisions for financial holding companies

This research analyzes the internationalization process model developed by Johanson and Vahlne and derives two integer programming investment decision models that consider the risk attitudes of investment firms. Johanson and Vahlne’s model provides a starting point for building a model that suits the investment approach and decision making process of financial holding companies. In practice, when firms make an international investment decision, there is a need for a model that can generate outputs based on financial measures such as profit, investment returns, and tolerable levels of risk. Thus, in this paper, Johanson and Vahlne’s concepts are studied and financial managers are interviewed to derive models that match the investment decision procedures of the firms. The model helps firms manage the risks of their investments and derive accurate investment strategies based on investment objectives and constraints.     

Log-linear modeling using conditional log-linear structures

Analysis of large dimensional contingency tables is rather difficult. Fienberg and Kim (1999, Journal of American Statistical Association, 94, 229–239) studied the problem of combining conditional (on single variable) log-linear structures for graphical models to obtain partial information about the full graphical log-linear model. In this paper, we consider the general log-linear models and obtain explicit representation for the log-linear parameters of the full model based on that of conditional structures. As a consequence, we give conditions under which a particular log-linear parameter is present or not in the full model. Some of the main results of Fienberg and Kim follow from our results. The explicit relationships between full model and the conditional structures are also presented. The connections between conditional structures and the layer structures are pointed out. We investigate also the hierarchical nature of the full model, based on conditional structures. Kim (2006, Computational Statistics and Data Analysis, 50, 2044–2064) analyzed graphical log-linear models based on conditional log-linear structures, when a set of variables is conditioned. For this case, we employ the Möbius inversion technique to obtain the interaction parameters of the full log-linear model, and discuss their properties. The hierarchical nature of the full model is also studied based on conditional structures. This result could be effectively used for the model selection also. As applications of our results, we have discussed several typical examples, including a real-life example.     

A Non-Parametric Approach to Model Evaluation

All existing tests and criteria for model evaluation are based on the sum of squared residuals. These measures do not evaluate the performance of a model in predicting turning points in the variable of interest. Here, three tests are proposed to evaluate models on the basis of their ability to predict turning points and to give an unbiased prediction of the relative order of magnitude of changes. Five models are evaluated using the proposed tests. Results indicate that the tests expose facets of models that R² and sum of squared residuals are unable to highlight. Hence, the tests are suggested to complement existing model evaluation techniques.     

On structures in instability zones

Two mathematical crystallization models describing structure formations in instability zones are proposed and justified. The first model, based on a phase field system, describes crystallization processes in binary alloys. The second model, based on a modified Biot model of a porous medium and the convective Cahn–Hilliard model, governs oriented crystallization. Physical interpretation and numerical analysis are discussed. Bibliography: 23 titles. Illustration: 5 figures.     

Visualizing Loglinear Models

Abstract

We consider visual methods based on mosaic plots for interpreting and modeling categorical data. Categorical data are most often modeled using loglinear models. For certain loglinear models, mosaic plots have unique shapes that do not depend on the actual data being modeled. These shapes reflect the structure of a model, defined by the presence and absence of particular model coefficients. Displaying the expected values of a loglinear model allows one to incorporate the residuals of the model graphically and to visually judge the adequacy of the loglinear fit. This procedure leads to stepwise interactive graphical modeling of loglinear models. We show that it often results in a deeper understanding of the structure of the data. Linking mosaic plots to other interactive displays offers additional power that allows the investigation of more complex dependence models than provided by static displays.     

Strong model dependence in statistical analysis: goodness of fit is not enough for model choice

Most statistical methods are based on models, but most practical applications ignore the fact that the results depend on the model as well as on the data. This paper examines the size of this model dependence, and finds that there can be very considerable variation between the results of fitting different models to the same data, even if the models being considered are restricted to those which give an acceptable fit to the data. Under reasonable regularity conditions, we show that different empirically acceptable models can give rise to non-overlapping confidence intervals for the same parameter. Application papers need to recognize that the validity of conventional statistical results rests on the assumption that the underlying model is known to be correct, and that this is a much stronger requirement than merely confirming that the model gives a good fit to the data. The problem of model dependence is only partially resolved by using formal methods of model selection or model averaging.

     

Granger因果检验的非线性进展及应用研究

Granger因果检验是计量经济学的重要组成部分,也是现代经济、金融学分析的重要工具,近年来Granger因果检验在非线性检验方向有了较大进展。本文在线性Granger因果检验的基础上,阐述了Granger因果检验的非线性进展,重点总结了针对一阶矩的基于回归模型、非参函数和信息理论的三大类非线性方法以及针对二阶矩的基于残差交叉相关系数和多元条件方差模型下的两大类非线性方法,讨论了不同非线性Granger方法中数据要求、核心模型、建模关键以及模型优缺点,提出了Granger因果检验"线性-非线性"的整体框架和研究范式.通过模型分析和比较,本文可为因果检验的非线性理论和模型研究提供参考,并对因果检验在经济和金融领域的更广泛应用提供支持。