AD HOC MODEL BUILDING USING TIME-VARYING PARAMETER MODELS*

Decision support for managers and policy makers, such as required in planning and evaluation efforts, often requires ad hoc behavioral modeling to account for context-specific phenomena and to handle data limitations. This paper introduces a systematic approach useful for meeting these requirements in which time-varying parameter estimation plays an important role. A case study evaluating the impacts of public policy actions on residential natural gas conservation illustrates the approach.     

How performance management regulations shape evaluation practice in South African municipalities

South African municipalities are taking hesitant steps towards institutionalising evaluation practice. This paper explores how evaluation is being institutionalised at a municipal level, looking at both the legislative and regulatory requirements in terms of planning, budgeting, and reporting, as well as opportunities for evaluative practice. A well-established performance management system, coupled with certain constraints to innovation and building an evaluative culture are contributing to the creation of an evaluation function that serves a monitoring purpose. A combination of desktop research and participatory action research was used to understand the municipal practice of evaluation. The results have implications for how decentralised public sector evaluation systems are understood.     

一种基于区间直觉梯形模糊数偏好的大群体决策冲突测度研究

针对决策偏好为区间直觉梯形模糊数的大群体决策冲突测度问题,首先给出了两个区间直觉梯形模糊数的距离,提出了群体成员冲突函数,基于该函数对大群体成员偏好进行聚类形成若干聚集,以此为基础提出了聚集冲突测度模型并集结为群体冲突测度模型,然后应用于大群体偏好集结。最后通过一个算例对模型进行了验证。该模型便于大群体意见的分析与协调,适用于群决策支持系统,在应急决策中具有实际应用价值。     

模糊车辆路径问题的一种混合遗传算法

在对模糊车辆路径问题进行简单描述的基础上,通过引入决策者主观偏好值的概念,给出了解决该问题的基本思路,建立了具有模糊特征的车辆路径问题的模糊机会规划模型,提出了求解该问题的一种基于模糊模拟的混合遗传算法。同时,在最小化总行驶距离的目标下,通过随机模拟方法研究了决策者主观偏好值的选择对最终决策目标的影响作用,并给出了其最佳取值范围。     

Causal factor relations of supply chain competitiveness via fuzzy DEMATEL method for Thai automotive industry

Abstract

The automotive industry is complex and global. An automotive supply chain involves multiple manufacturers and service providers in different parts of the chain across different countries. Diffusion of the global production system creates both opportunities and challenges for the supply chain itself as well as production base countries. Thailand is a production base for over 15 automotive brands. The industry contributes significantly to the Thai economy and employs approximately half a million people. To address competitiveness of the Thai automotive industry, its supply chain is studied in this paper. Ten factors affecting the competitiveness of the Thai automotive supply chain are analysed through the fuzzy DEMATEL method. The method forms a structural model that categorises the causal or effect roles of the factors, and the degree at which they impact the supply chain. This study focuses on Tier 1 and Tier 2 manufacturers of the supply chain. The results reveal that there are three common causal factors and four common effects in both tiers. The other four factors play different roles in different tiers. Managerial implications relating to these factors are also discussed and certain recommendations proposed in order to enhance the competitiveness of the Thai automotive supply chain.     

Multiplicative consistency analysis for interval fuzzy preference relations: A comparative study

To examine the multiplicative consistency of interval fuzzy preference relations (IFPRs), this paper first analyzes the limitations associated with the previous consistency concepts. Accordingly, a new consistency concept is defined that is an extension of the crisp case and overcomes limitations in the previous concepts. Next, a linear programming model to judge the consistency of IFPRs is constructed, and an approach to derive multiplicative consistent IFPRs is introduced. Furthermore, goal-programming models to determine missing values in an incomplete IFPR are constructed that have the highest consistent level with respect to known values. Moreover, a multiplicative consistency and consensus based method for group decision making with IFPRs is developed that can address incomplete and inconsistent cases. Finally, two practical decision-making problems are offered to demonstrate the feasibility and efficiency of the new method, and an analysis of a numerical and theoretical comparison with several related methods is performed.     

Erfolgsfaktoren beim Einzel-Coaching

Success factors of executive coaching. A screening of the coaching scenery from the client’s point of viewThis survey, based on questionnaires, examines the success factors of one-to-one coaching from the client point of view. Firstly, ten success factors in coaching were established by means of factor analysis. Secondly, more extensive analysis shows that, contrary to previous research in this field, the “professional system” is not necessarily a determining success factor in coaching. As an exclusive instrument of staff development, coaching seems to be effective even without support from the professional system. The analysis also indicates that the clients perceive their coach as an advisor and confidant as regards role expectations. Furthermore, multiple regression analysis discloses the predictive effects of the two success factors “the coach’s qualification” and “the coach’s involvement” for the overall success of coaching from the client’s point of view. The overall success of coaching itself is established through three items: the clients’ attitudes towards coaching, their attitude concerning further coaching, and the extent to which their objectives were met in coaching.     

基于犹豫直觉模糊语言集均值-标准差偏好距离的多属性决策研究

在犹豫直觉模糊语言集和语言尺度函数的基础上定义了均值-标准差偏好的Hamming距离，并提出了基于犹豫直觉模糊语言集距离TOPSIS和TODIM的多属性决策方法，进一步利用这两种方法对实例问题建筑商的招标方案进行排序，并讨论了偏好参数对排序结果的灵敏度分析。为验证上述方法的有效性，与已有方法对同一实例的排序结果进行了比较分析。     

基于HTFWGBM算子的供应商选择模型研究

由于供应商选择中的的复杂性与不确定性以及人类认知的有限性,从而导致信息融合失真和决策结果不准确的情况。针对决策属性以犹豫三角模糊数（HTFN）给出的供应商的选择问题,本文提出一种基于HTFGWBM算子的决策算法。首先,针对犹豫三角模糊数和几何Bonferroni平均算子理论,分别定义了犹豫三角模糊几何Bonferroni平均（HTFGBM）算子和犹豫三角模糊几何加权Bonferroni平均（HTFGWBM）算子,同时分别研究了算子的幂等性,置换不变性,单调性和有界性等性质。其次,基于HTFGWBM算子构建新型犹豫多属性决策模型,结合HTFN排序方法进行备选供应商排序。最后通过算南水北调中线工程中的供应商选择实例证明了决策模型的可行性与有效性。结果表明,通过调整模型参数,模型具有一定的延展性和容错能力,能够很好的进行科学决策。