基于残缺值的模糊偏好关系多属性大群体决策方法

在多属性大群体决策环境下，针对决策者给出关于决策方案两两比较且可能包含残缺值的模糊偏好关系的决策问题，提出了一种基于相似度聚类的残缺值模糊偏好关系大群体决策方法。此方法首先通过标准化残缺值矩阵，然后定义判断矩阵之间的相似度对大群体进行聚类，再对各个属性下的群体偏好进行集结，通过偏差熵模型确定各个决策属性的权重，集结所有决策属性下的群体偏好，最后得到决策方案的排序结果。文章最后给出了一个算例分析以验证此方法的有效性。     

一种基于广义相似度的物资紧急调运方案优选方法

物资紧急调运方案优选决策目标具有不可公度和相互矛盾性,其属性值既有实数、区间数,也有语言值和不确定语言值等,无法统一度量。在运用TOPSIS方法来确定理想方案的基础上,通过定义一种广义相似度,将备选方案的各类型属性值的相似度计算规范到统一的度量空间内,实现了以实数、区间数、语言值和不确定语言值表示的属性值的统一度量。而后,提出了备选方案与理想方案的综合关联度,并以此来对备选方案排序,初步解决了物资紧急调运方案优选这一混合型多属性决策问题。最后,给出了一个实例说明了该方法具有良好的适应性。     

民办医学高校兼职教师的选择研究

目的:分析在各个属性评价信息不能完全确知的情况下,民办医学高校管理者科学的选择兼职教师。方法:考虑到决策者本身存在一定主观偏好的现象,作为被选择的兼职教师本身,也受到原单位工作性质和工作环境不确定性的限制,不能给出确定的属性值。采用不确定多属性决策模型进行分析研究,并引进多目标决策的交互式决策思想。结果:结合决策者的主观偏好对所确定的决策方案集进行缩减,最终求得满意解。文章最后给出了具体算例。     

多属性决策方法及其在节能减排中的应用研究

通过对直觉模糊数的区间化处理,把直觉模糊多属性决策转化为属性值为区间数的多属性决策.基于决策者的主观偏好值和客观偏好值的总偏差最小化原则得到属性权重,从而给出了一种直觉模糊多属性决策方法.最后利用所提出的方法对企业的节能减排效果进行了评价研究.     

一种基于梯形模糊中心的不确定多属性决策Topsis方法

文章针对只有部分属性权重信息且属性值以梯形模糊数形式给出的模糊多属性决策问题,提出一种基于梯形模糊数的中心的多属性决策方法。该方法建立一个非线性规划模型,通过求解该模型得到属性的权重向量。然后根据传统的Topsis方法,计算各方案属性值加权后的模糊中心与正负理想中心的贴近度的大小进行排序择优,最后给出算例。     

一种基于模糊距离和证据理论的多属性群决策方法

针对语言评价信息形式的多属性群决策问题,提出了一种基于模糊距离和D-S证据理论的群决策方法。该方法首先运用基于模糊距离的方法来获得决策者权重,再对多个决策者给出的语言评价信息进行分析,然后通过D-S合成法则对其值进行合成。计算信度函数和似真函数,据此对所有决策方案进行排序。最后通过一个算例验证该方法的可行性和有效性。     

考虑属性关联的不确定语言变量R-TOPSIS模型

采用不确定语言变量进行决策是有效反映客观事物的复杂不确定性以及人类思维的模糊性手段。针对多属性决策问题中属性之间具有相互关联作用这一问题,在引入加权决策距离矩阵和Choquet积分构建了一种考虑属性关联的R-TOPSIS语言决策方法。最后结合案例说明该方法的决策过程以及该方法的可行性和有效性。     

概率有序加权算子P-OWG在公路方案优选中的应用

在公路方案优选的多属性方案决策中,运用概率有序加权几何(P-OWG)算子对各方案的属性值进行集成,根据得出的综合属性值对各方案进行排序和优选,可以十分方便的选出最优的方案。     

应用加权函数排序法求解模糊需求库存模型

应用代表决策者偏好的加权函数排序法对已有的不定需求单一周期库存模型进行求解。在之前对库存模型的讨论中,许多学者用不同的方法来求解模型,取得了许多成果。但存在着不足之处,即无法在模型中体现决策者的意志。在现实中,决策通常是由决策者主观决定的。因此,文章引入了决策偏好参数,并在Yager排序法的基础上添加了代表决策者偏好的加权函数对不确定需求的单一周期库存模型求解,以找到模糊情况下总成本的估计,从而推导出与之相对应的最优定购量。结合决策偏好参数对所求得的一一对应的最优解曲线进行解释。最后应用三角模糊数的实例对文章的结果进行讨论。     

基于区间数的大学生综合素质多属性群决策方法及应用

目的:针对大学生综合素质属性多样化、不可重复的特性,导致大学生综合素质专家对大学生综合素质评价模糊,决策信息不完备的问题,提出了一种基于区间数的群决策方法。方法:首先,构建了大学生综合素质综合评价指标体系,分析并得到属性指标的权重;结合区间数间距离建立非线性规划模型来确定专家权重;然后,再利用区间数算术加权平均算子将多个决策矩阵进行有效地集结,形成综合决策矩阵,结果:通过建立比较可能度矩阵,利用排序向量法对大学生综合素质评价决策矩阵进行择优排序。结论:实例表明,利用区间数的群决策方法对大学生综合素质进行评价,取得了良好的效果,为大学生综合素质评价这一类复杂的决策问题提供了有效的途径以及可借鉴的方法。     

Fuzzy multi-linguistic preferences model of service innovations at wholesale service delivery

In group decision making, most researches often assume the linguistic ways with personal preferences have been given and ignore the linguistic evaluated formats involving their knowledge and experience. In practice, people contributing to the judgment tend generally to give ratings about their personal preferences depending on their background. Thus, problems in multiple linguistic preferences go undetected, resulting in the evaluation process not satisfying with decisions’ expectations. In this study, we provide a fuzzy multiple preference integrated model with two stages to better reduce the bias for group decision makings. The first stage focuses on making the information unify on the alternatives according to the individual linguistic preferences, then we compute collective performance values and solve the problems lacking on the integration of respective fuzzy choice subsets. The second stage, we choose the alternatives of retailing service innovations according to the collective performance values by stage one. The goal of the decision process is to reach the subjective fuzzy cognitions in terms of the preference values of all the decision makers. Finally, the survey data of the chain wholesale using multiple preference formats in service innovations determination is verified.     

Adjusting inconsistency through learning in group decision-making, and its application to China's MBA recruiting interview

In group decision-making, because of limitations on individual knowledge and information bases, or because of the existing decision rule, an individual decision maker may not be capable of evaluating selected alternatives. Such circumstances can lead to inconsistencies across group decision matrices. These inconsistencies are difficult to remedy under existing approaches. Based on Rough Set Theory, we thus propose a new approach that integrates two types of learning techniques. It first applies a machine-learning procedure that extracts possible alternatives from other decision makers that are currently not included in a given decision maker's alternative set. It then applies a group knowledge-learning model to determine corresponding attribute values of those newly learned alternatives in meeting a group's consistency requirement. Efficacy of the approach is illustrated by its application to China's MBA recruiting interview.     

Shortlisting procedure with a limited capacity

A decision maker, who is overwhelmed by the number of available alternatives, limits her consideration. We investigate a model where a decision maker’s capacity determines whether she is overwhelmed: She considers all the available alternatives if their number does not exceed her capacity; otherwise, she applies a shortlisting procedure to reduce the number of alternatives to within her capacity. We show how to deduce the decision maker’s capacity, her preference and the alternatives that she considers from the observed behavior. Furthermore, we provide the necessary and sufficient conditions for a consideration function to be derived by the shortlisting procedure with a limited capacity.     

A model of stochastic choice from lists

We study stochastic choice from lists. All lists present the same set of alternatives albeit in different orders. Faced with a list, the decision maker makes her choice in two stages. In the first stage she searches through the list till she sees $k$ alternatives. In the second stage she chooses from the alternatives she has seen. Both $k$ and the choice rule governing her second stage behavior are random. We show that the underlying primitives of our model are revealed by the decision maker’s choice frequencies from lists. We characterize the model and two of its special cases. In the first special case the decision maker deterministically chooses the best observed alternative according to a given preference. In the second, the decision maker maximizes random preferences.     

Similarity-based mistakes in choice

We characterize the following choice procedure. The decision maker is endowed with two binary relations over alternatives, a preference and a similarity. In every choice problem she includes in her choice set all alternatives which are similar to the best feasible alternative. Hence she can, by mistake, choose an inferior option because it is similar to the best. We characterize this boundedly rational behavior by suitably weakening the rationalizability axiom of Arrow (1959). We also characterize a variation where the decision maker chooses alternatives on the basis of their similarities to attractive yet infeasible options. We show that similarity-based mistakes of either kind lead to cyclical behavior. Finally, we reinterpret our procedure as a method for choosing a bundle given a set of individual items, in which the decision maker combines the best feasible item with those that complement it.     

基于多粒度二元语义的供应商选择方法

供应商选择是一个多属性决策问题,既包括定性的因素也包括定量的因素。把成本、质量、服务水平、供应商的能力和外部因素作为供应商选择的标准,并将基于多粒度二元语义信息处理的群决策方法应用于供应商的选择,给出了基于二元语义信息处理的群决策的步骤,最后给出了一个算例。     

The assessment of the expected value of information in the binary decision model

The expected value of information represents the maximum amount the decision maker should spend on inquiry before making a decision. This amount depends upon the accuracy of the information. In many cases of inquiry, prior objective knowledge of the accuracy is not available. This paper presents and compares two methods of subjectively assessing the value of imperfect information in the binary decision model. In the first method, the decision maker provides a likelihood function for the inquiry and hence the probabilities of error. The second method is the preposterior approach, in which the decision maker provides the prior distribution for the posterior probability.     

A theory of iterative choice in lists

In a list, alternatives appear according to an order and the decision maker follows this order to evaluate alternatives. He records the first alternative as the initial survivor and then at every stage, he compares the current survivor with the next alternative in the list to determine whether the next alternative replaces that to become the new survivor. When the entire list is exhausted in this manner, the agent chooses the survivor in the last stage. We call this procedure “iterative” and provide an axiomatic characterization for it when the order in every list is observable. Then, we also study characterizations of the iterative procedure that is prone to the well-known primacy and recency effects. Finally, we analyze situations where the order of alternatives is unknown to an outside observer and provide a characterization result that enables such an outsider with limited information to understand whether the decision maker can indeed be an iterative list chooser for some order.     

Decision making with imprecise probabilistic information

We develop an axiomatic approach to decision under uncertainty that explicitly takes into account the information available to the decision maker. The information is described by a set of priors and a reference prior. We define a notion of imprecision for this informational setting and show that a decision maker who is averse to information imprecision maximizes the minimum expected utility computed with respect to a subset of the set of initially given priors. The extent to which this set is reduced can be seen as a measure of imprecision aversion. This approach thus allows a lot of flexibility in modelling the decision maker attitude towards imprecision. In contrast, applying Gilboa and Schmeidler [J. Math. Econ. 18 (1989) 141] maxmin criterion to the initial set of priors amounts to assuming extreme pessimism.