基于集成支持向量机的企业财务业绩分类模型研究

要想正确预测公司财务业绩,首先必须选择合适的预测方法。现有文献所采用的财务业绩预测模型普遍存在着泛化能力不强的问题。本文提出用支持向量机方法来预测我国上市公司的财务业绩。为了提高预测准确率,本文还用AdaBoost算法对支持向量机进行了改进(集成支持向量机)。在支持向量机核函数的选择上,我们采用了实验法,即对每个核函数及其相关参数的预测效果都进行了测算,以期找出最适用的预测模型。实证结果表明,径向基核函数(rbf)的效果最好,支持向量机方法预测准确率远远高于其它方法。     

Testing Prospect Theory in a Deterministic Multiple Criteria Decision-Making Environment*

Prospect theory by Kahneman and Tversky [7] is tested in a deterministic multiple criteria decision-making context. In two experiments conducted in classroom settings subjects made pairwise preference comparisons of condominiums for sale. The results of the experiments indicate that the traditional value model did not explain the subjects' revealed preferences as well as the prospect model. We conclude that prospect theory is a reasonable model of choice for many individuals in such a context.     

Shopping Model of the Time-Sensitive Consumer

The shopping behavior of time-sensitive consumers has been modeled as an economic model of choice. Consumers are said to balance the costs of time spent commuting to the store and in the store with storage costs and other nontime costs of shopping. In so doing, these consumers tend to minimize their overall costs. Propositions are developed and tested using this economic model of shopping. The model is extended beyond time sensitivity to include deal proneness; further propositions are developed and tested using the extended model. The empirical results support the derived models.     

EXPLORING MARKET PRICING STRATEGIES VIA DYNAMIC PROGRAMMING

Price determinants as well as strategies can be studies by use of simulation, particularly if cost and price relationships can be related to market activity [1] [9] [11]. But, through the use of dynamic programming, given the market conditions, one can extend the analysis to include an optimal strategy. This paper describes a dynamic programming approach to studying price strategy. A model is developed to show that in a market characterized by cost/volume and price/volume relationships, profitability can be extended beyond that resulting from a dominant market strategy to an optimal maximizing strategy. Extension of the model is suggested for studying (a) sensitivity of a strategy (solution) to price level and cost changes, (b) optimal timing of withdrawal, and (c) present value analysis.     

DEA models for extended two-stage network structures

Liang et al. (2008) [1] developed DEA models based upon game approach to decompose efficiency for two-stage network structures where all outputs of the first stage are the only inputs to the second stage. This paper extends Liang et al. (2008) [1] by assuming that the inputs to the second stage include both the outputs from the first stage and additional inputs to the second stage. Two models are proposed to evaluate the performance of this type general two-stage network structures. One is a non-linear centralized model whose global optimal solutions can be estimated using a heuristic search procedure. The other is a non-cooperative model, in which one of the stages is regarded as the leader and the other is the follower. The newly developed models are applied to a case of regional R&D of China.     

Learning to learn and productivity growth: Evidence from a new car-assembly plant

This paper models learning by experience beyond the experience curve, including the possibility of “learning to learn”: the pace of learning increases over time by building on what has already been learned. We compare the extended deterministic learning model with Jovanovic and Nyarkos' [26] stochastic learning. The theoretical models are tested with data on the total factor productivity of a car-assembly plant in its first months of operation. We find that the deterministic “mixed learning model”, where the speed of learning is equal to a constant plus a learning to learn effect, is the one that best fits the empirical data. The mixed learning model results in a time pattern of total factor productivity growth, first increasing and later decreasing, different from the always decreasing rate of growth of the learning curve, opening new perspectives on the study of learning by experience.     

Spatial Transmission Models: A Taxonomy and Framework

Within risk analysis and, more broadly, the decision behind the choice of which modeling technique to use to study the spread of disease, epidemics, fires, technology, rumors, or, more generally, spatial dynamics, is not well documented. While individual models are well defined and the modeling techniques are well understood by practitioners, there is little deliberate choice made as to the type of model to be used, with modelers using techniques that are well accepted in the field, sometimes with little thought as to whether alternative modeling techniques could or should be used. In this article, we divide modeling techniques for spatial transmission into four main categories: population‐level models, where a macro‐level estimate of the infected population is required; cellular models, where the transmission takes place between connected domains, but is restricted to a fixed topology of neighboring cells; network models, where host‐to‐host transmission routes are modeled, either as planar spatial graphs or where shortcuts can take place as in social networks; and, finally, agent‐based models that model the local transmission between agents, either as host‐to‐host geographical contacts, or by modeling the movement of the disease vector, with dynamic movement of hosts and vectors possible, on a Euclidian space or a more complex space deformed by the existence of information about the topology of the landscape. We summarize these techniques by introducing a taxonomy classifying these modeling approaches. Finally, we present a framework for choosing the most appropriate spatial modeling method, highlighting the links between seemingly disparate methodologies, bearing in mind that the choice of technique rests with the subject expert.     

Some Comments on the Validity of EOQ Formula under Inflationary Conditions

In an earlier issue of Decision Sciences, Jesse, Mitra, and Cox [1] examined the impact of inflationary conditions on the economic order quantity (EOQ) formula. Specifically, the authors analyzed the effect of inflation on order quantity decisions by means of a model that takes into account both inflationary trends and time discounting (over an infinite time horizon). In their analysis, the authors utilized two models: Current-dollars model and Constant-dollars model. These models were derived, of course, by setting up a total cost equation in the usual manner then finding the optimum order quantity that minimizes the total cost. Jesse, Mitra, and Cox [1] found that EOQ is approximately the same under both conditions; with or without inflation. However, we disagree with the conclusion drawn by [2] and show that EOQ will be different under inflationary conditions, provided that the inflationary conditions are properly accounted for in the formulation of the total cost model.     

MNC–SME strategic alliances — A model framing knowledge value as the primary predictor of governance modal choice

This paper addresses a deficit in research on strategic alliance formation which, as noted by Gulati (1998) [Gulati, R., 1998. Alliances and networks. Strategic Management Journal 19, 293–317], has resulted in little attention being given to transaction costs and knowledge-based imperatives for alliance formation. In addition, Contractor and Ra (2002) [Contractor, F.J., Ra, W., 2002. How knowledge attributes influence alliance governance choices: a theory development note. Journal of International Management 8, 11–27] illustrated that theory on governance modal choice from the perspective of transactions costs and knowledge characteristics is underdeveloped, thereby proposing an interactive model predicting governance modal choice in an attempt to redress this imbalance. Using a multiple case study methodology the present study examines the important aspects of governance mode and knowledge exchange in strategic alliances between a multi-national corporation (MNC) and small and medium enterprises (SMEs). In doing so this paper qualifies the model proposed by Contractor and Ra (2002) [Contractor, F.J., Ra, W., 2002. How knowledge attributes influence alliance governance choices: a theory development note. Journal of International Management 8, 11–27], suggesting a model framing knowledge value as the primary predictor of governance modal choice as an appropriate framework to evaluate both the knowledge exchange relationships and governance modal choice relationships between MNCs and SMEs. Thus this paper advances theory on the motives for alliances, specifically, the effect on governance modal choice of a company's desire to acquire or access knowledge. In addition the paper furthers theory on organisational learning and absorptive capacity focusing on their influence on governance mode, in the context of management strategy.     

INDUSTRY RESOURCE ALLOCATION TO BASIC RESEARCH UNDER NORMALLY DISTRIBUTED BENEFITS*

The failure of a free market system to attain socially optimal allocation of resources to research and development (R&D) is a generally recognized problem. However, we are just beginning to understand the types of R&D activities that receive relatively serious underinvestment from specific types of industries and the types of governmental intervention strategies that are likely to be effective and efficient in the correction of that underinvestment. Recently, Joglekar and Hamburg [16] [17] sought answers to these types of questions using models of the resource allocation behavior of firms considering investment in basic research related to their industry. It was assumed that the firms' benefits were exponentially distributed. In the present article, such benefits are assumed to be normally distributed, and an attempt is made to verify the earlier conclusions and policy implications of [16] and [17]. The results are similar for these two substantially different types of distribution, but the earlier conclusions and policy implications are clarified, qualified, and extended.     

SOME ASPECTS OF THE DYNAMIC PROPERTIES OF ECONOMETRIC MODELS*

Recent efforts by several authors have resulted in the development of techniques for analytically describing the dynamic properties of linear stochastic difference equation models ¹ 1 For examples of this work see Chow [1]; Chow and Levitan [2] [3]; Howrey [8] [9]; Howrey and Kelejian [10] and Howrey and Klein [11].
. Since these techniques have not been extensively applied to econometric models and consequently are not generally familiar to economic modelers, the purpose of this paper is to summarize the techniques these researchers have developed and to illustrate them using a simple (and admittedly unrealistic) econometric model.     

THE ANALYSIS OF OUTLYING DATA POINTS USING ROBUST REGRESSION: A MULTIVARIATE PROBLEM-BANK IDENTIFICATION MODEL*

Because the eight largest bank failures in United States history have occurred since 1973 [24], the development of early-warning problem-bank identification models is an important undertaking. It has been shown previously [3] [5] that M-estimator robust regression provides such a model. The present paper develops a similar model for the multivariate case using both a robustified Mahalanobis distance analysis [21] and principal components analysis [10]. In addition to providing a successful presumptive problem-bank identification model, combining the use of the M-estimator robust regression procedure and the robust Mahalanobis distance procedure with principal components analysis is also demonstrated to be a general method of outlier detection. The results from using these procedures are compared to some previously suggested procedures, and general conclusions are drawn.     

An enhanced BAM for unbounded or partially bounded CRS additive models

The Bounded Adjusted Measure (BAM), initially defined for the additive model, which is a variable returns to scale (VRS) model, was extended to the constant returns to scale (CRS) case [7]. The added range-bounds, which maintain unaltered the production possibility set (PPS) under VRS, showed an influential effect under CRS, reducing the corresponding PPS, as well as a negative effect, excluding some of the original CRS projections. Here we propose an enhanced extension that, by considering a different set of less restrictive bounds, eliminates the negative effect. Moreover, we customize this new extension for the family of partially bounded CRS additive models, i.e., models where at least one variable is naturally bounded from below, if it is an input, or from above, if it is an output.     

Sharp Identification Regions in Models With Convex Moment Predictions

We provide a tractable characterization of the sharp identification region of the parameter vector θ in a broad class of incomplete econometric models. Models in this class have set‐valued predictions that yield a convex set of conditional or unconditional moments for the observable model variables. In short, we call these models with convex moment predictions. Examples include static, simultaneous‐move finite games of complete and incomplete information in the presence of multiple equilibria; best linear predictors with interval outcome and covariate data; and random utility models of multinomial choice in the presence of interval regressors data. Given a candidate value for θ, we establish that the convex set of moments yielded by the model predictions can be represented as the Aumann expectation of a properly defined random set. The sharp identification region of θ, denoted Θ_I, can then be obtained as the set of minimizers of the distance from a properly specified vector of moments of random variables to this Aumann expectation. Algorithms in convex programming can be exploited to efficiently verify whether a candidate θ is in Θ_I. We use examples analyzed in the literature to illustrate the gains in identification and computational tractability afforded by our method.     

Note on solving inventory models from operational research view

In this article, we will show that an optimal solution can be derived without complicated mathematical derivation, an improvement on the paper [Yang, G.K., et al., 2011. Note on inventory models with Weibull distribution deterioration. Production Planning & Control, 22 (4), 437–444] that was recently published in Production Planning & Control. We extend their inventory model with a generalised deterioration and backlog rates. Our findings will allow researchers to consider inventory models directly from the view of operational research instead of complex analysis.     

When are two multivariate random processes indistinguishable

Prediction error identification methods have been recently the objects of much study, and have wide applicability. The maximum likelihood (ML) identification methods for Gaussian models and the least squares prediction error method (LSPE) are special cases of the general approach. In this paper, we investigate conditions for distinguishability or identifiability of multivariate random processes, for both continuous and discrete observation time T. We consider stationary stochastic processes, for the ML and LSPE methods, and for large observation interval T, we resolve the identifiability question. Our analysis begins by considering stationary autoregressive moving average models, but the conclusions apply for general stationary, stable vector models. The limiting value for T → ∞ of the criterion function is evaluated, and it is viewed as a distance measure in the parameter space of the model. The main new result of this paper is to specify the equivalence classes of stationary models that achieve the global minimization of the above distance measure, and hence to determine precisely the classes of models that are not identifiable from each other. The new conclusions are useful for parameterizing multivariate stationary models in system identification problems. Relationships to previously discovered identifiability conditions are discussed.     

DYNAMIC DISAGGREGATE CHOICE MODELS,WITH AN APPLICATION IN TRANSPORTATION

A standard assumption of disaggregate mode-choice models is that an individual faced with a choice among several available alternative transportation modes compares their utilities. These utilities are almost always expressed as a linear composite measure of various perceived benefits which are assumed constant. However, the individual's perceptions change as he learns, acquires new habits, or adapts to different physical, economic, and social circumstances. In order to account for these changes, two dynamic models of model-choice behavior are developed herein. These two models are based on two common postulates. (1) One alternative is preferred over another only if the absolute difference in their utilities exceeds a positive constant; otherwise, the decision maker is indifferent toward the two alternatives. (2) If an alternative is preferred, it will be chosen with certainty. In the indifference state, the individual is postulated to randomly choose one of the two alternatives (Model 1) or choose the same alternative as was most recently chosen (Model 2). Choice probabilities derived from these two models are shown to differ from those obtained using the conventional logit model. If there is a strong loyalty toward a mode, the logit model underestimates its choice probability when that mode is less attractive than the competing mode. The results are illustrated using numerical examples.     

拓展的NBD-Dirichlet模型及其在零售店选择研究中的应用

纯随机NBD-Dirichlet模型可以用来研究消费者的零售店选择行为(商店选择与购买频次),但是由于模型中没有协变量,所以无法对消费者的商店选择行为进行解释.本论文对纯随机模型进行了拓展,引入了解释变量,并应用拓展的模型对北京的消费者超市选择行为进行了实证研究.其中的主要发现为:(1)市场渗透率是超市经营的关键指标;(2)促销没有对消费者超市选择产生影响,但是它能够使顾客更多地惠顾超市;(3)家庭收入对消费者选择超市以及惠顾频次的多少均有显著的影响;(4)单次购买金额对消费者选择超市没有显著影响,但是它对消费者惠顾频次有负向的影响;(5)超市之间的价格差异较小,因此价格对消费者购物频次没有影响;(6)产品的齐全性对消费者超市选择有巨大的影响.     

时变参数模型的最优滚动窗宽选择标准及应用

宏观经济领域中存在严重的结构突变性,模型估计量的优劣对估计样本规模是敏感的。本文针对时变参数模型,建立了滚动窗宽选择标准,通过最小化估计量的近似二次损失函数及最大化各子样本估计量间的曼哈顿距离选择窗宽大小,权衡了模型估计量的准确性和时变性两个相悖目标。蒙特卡罗模拟实验表明,本文所提出的方法在各种结构突变情形下均适用,能够应用于线性关系和非线性关系的时变参数模型中,且均具有稳健性。将该方法应用于我国金融网络的结构突变识别过程,显著改善了传统窗宽选择方法的结果。