A priori identification of preferred alternatives of OWA operators by relational analysis of arguments

We present a reverse decision-aiding method that is distinct from previously reported ordered weighted averaging (OWA) aggregation methods. The proposed method is implemented in two phases. In the first phase, potentially best alternative, defined as one having any weighting vector that enables it to be at least as good as the others, is identified. In the second phase, the maximum and the minimum attitudinal characters for such an alternative are computed as the highest and the lowest values which it can attain under the weights-set identified in the first phase. These two phases are governed only by the relational analysis of input arguments, without soliciting the decision-maker to supply a specific attitudinal character. The proposed method can be applied to cases when it is difficult to obtain a precise attitudinal character and when, even if a precise one is obtained, the OWA operator weights are different, depending on the weights generating methods adopted. Further, if uncertain attitudinal character in the form of the interval number is available, its projection into the results of the proposed method yields less alternatives of consideration, in some cases, a single best alternative. The proposed method also allows for a priori identification of alternatives prone to change at a particular range of attitudinal character.     

Implicit Elicitation of Attitudinal Character in the OWA Operator

Ordered weighted averaging (OWA) operators are well recognized as aggregation operators and are successfully applied to real world problems. The OWA operator weights are also known to play an important role in the aggregation, and thus, they heavily affect the final aggregation results. In this paper, we attempt to elicit a hidden attitudinal character from a decision‐maker through a simple question about his or her preference of alternatives for the purpose of alleviating the burden of specifying an attitudinal character. A paired comparison of alternatives in a decision‐making problem yields a precise (or uncertain) attitudinal character, which can be used to derive OWA operator weights via one of the well‐established weight generating methods.     

Rule extraction from support vector machines by genetic algorithms

Support vector machines (SVMs) are state-of-the-art tools used to address issues pertinent to classification. However, the explanation capabilities of SVMs are also their main weakness, which is why SVMs are typically regarded as incomprehensible black box models. In the present study, a rule extraction algorithm to extract the comprehensible rule from SVMs and enhance their explanation capability is proposed. The proposed algorithm seeks to use the support vectors from a training model of SVMs and combine genetic algorithms for constructing rule sets. The proposed method can not only generate rule sets from SVMs based on the mixed discrete and continuous variables but can also select important variables in the rule set simultaneously. Measurements of accuracy, sensitivity, specificity, and fidelity are utilized to compare the performance of the proposed method with direct learner algorithms and several rule-extraction techniques from SVMs. The results indicate that the proposed method performs at least as well as with the most successful direct rule learners. Finally, an actual case of pressure ulcer was studied, and the results indicated the practicality of our proposed method in real applications.     

Determination of the Attitudinal Character by Self‐Evaluation for the Maximum Entropy OWA Approach

The attitudinal character plays an important role in the maximum entropy ordered weighted averaging (MEOWA) approach. We propose a self‐evaluation model to determine the attitudinal character for each alternative in multicriteria decision making with MEOWA operators. The value of the attitudinal character determined for each alternative by self‐evaluation may be different; and each alternative can reach its highest rank with MEOWA weights, when the attitudinal character is determined for it. Then, to obtain an overall set of MEOWA weights by different attitudinal character values, the preemptive goal programming (PGP) model proposed by Wang can be used. Finally, the evaluation process of the service quality for five online stores is given to illustrate the application of the proposed method.     

判定真实谱峰位置的一个通用准则

提出了许多改善ＦＦＴ离散谱估计精度的方法,但必需知道谱峰的大体范围。作者证明了对称窗函数的频谱主瓣内的两条ＦＦＴ谱线为最大和次大。根据这一性质,对多频率信号,可按照ＦＦＴ谱线局部极大的原则确定真实谱峰位置,这是非常简便且易于编程的准则。仿真结果表明,如果两个真实谱峰相距不少于三条谱线,上述准则是正确的;随着谱峰接近,则可能存在误判,并且加汉宁窗错误率低于矩形窗。     

Behavioral,data-driven,agent-based evacuation simulation for building safety design using machine learning and discrete choice models

To improve occupant safety during building emergencies, evacuation simulations have been widely used for building safety design. Since occupant behavior is a determining factor for the outcome of building emergencies, accurately capturing how occupants make decisions and integrating occupants’ decision-making processes in evacuation simulations is important. In this study, based on the results of fire evacuation experiments in a virtual metro station, how different social (crowd flow) and environmental (visual access and vertical movement) factors would affect individuals’ wayfinding behavior was predicted using machine learning and discrete choice models. The trained models were further employed in agent-based evacuation simulations to examine crowd evacuation performance under different building design scenarios. Both the machine learning and discrete choice models could accurately predict individuals’ directional choices during emergency evacuations. Different building attributes could collectively influence occupant behavior, leading to distinct exit choices and evacuation times. While both the trained machine learning and discrete choice models generated similar results, the discrete choice model had better interpretability. Moreover, by comparing the trained models in this study with a model developed in a prior study, it was found that agents had significantly distinct responses to different building designs. Critical factors (e.g., type and size of buildings, occupants’ familiarity with the building) for the applicability of evacuation models were identified. Furthermore, recommendations were provided for future research that aims at employing evacuation simulations for building design evaluation and optimization.     

A note comparing support vector machines and ordered choice models’ predictions of international banks’ ratings

We find that support vector machines can produce notably better predictions of international bank ratings than the standard method currently used for this purpose, ordered choice models. This appears due to the support vector machine's ability to estimate a large number of country dummies unrestrictedly, which was not possible with the ordered choice models due to the low sample size.     

Discrete choice analysis of spatial attack sites

This paper presents an algorithm for the complete specification of multinomial discrete choice models to predict the spatial preferences of attackers. The formulation employed is a modification of models previously applied in transportation flow and crime analysis. A breaking and entering crime data set is employed to compare the efficacy of this model with traditional hot spot models. Discrete choice models are shown to perform as well as, or better than such models and offer more interpretable results.

Modelling human perception of static facial expressions

A recent internet based survey of over 35,000 samples has shown that when different human observers are asked to assign labels to static human facial expressions, different individuals categorize differently the same image. This fact results in a lack of an unique ground-truth, an assumption held by the large majority of existing models for classification. This is especially true for highly ambiguous expressions, especially in the lack of a dynamic context. In this paper we propose to address this shortcoming by the use of discrete choice models (DCM) to describe the choice a human observer is faced to when assigning labels to static facial expressions. Different models of increasing complexity are specified to capture the causal effect between features of an image and its associated expression, using several combinations of different measurements. The sets of measurements we used are largely inspired by FACS but also introduce some new ideas, specific to a static framework. These models are calibrated using maximum likelihood techniques and they are compared with each other using a likelihood ratio test, in order to test for significance in the improvement resulting from adding supplemental features. Through a cross-validation procedure we assess the validity of our approach against overfitting and we provide a comparison with an alternative model based on Neural Networks for benchmark purposes.     

Reference-dependent preferences on smart phones in South Korea: Focusing on attributes with heterogeneous preference direction

Consumers observe reference-dependent comparative attribute levels rather than absolute attribute levels of a product in their actual purchasing process. As the reference-dependent effect is important, many consumer preference studies in psychology and behavioral economics have perceived the reference-dependence as a general assumption. Nonetheless, few studies utilize standard economic models such as the discrete choice model, which is useful for analyzing demand forecasting in the marketing sector, and have used a reference-dependence utility function. Moreover, these studies analyze attributes only with the identical consumer preference directions, such as time and cost. Therefore, this study proposes an advanced discrete choice model that analyzes asymmetry preferences for attributes with heterogeneous preference direction as well as reference-dependence effect. It also investigates empirically the usefulness of the model on the market for smart phones. Based on this approach, we find wide disparities in consumer asymmetry preferences and loss aversion parameters in accordance to the existence of market experience and product attributes.     

Generalized attitudinal Choquet integral

Attitudinal Choquet integral (ACI) extends Choquet integral (CI) through a consideration of a decision-maker's (DM's) attitudinal character. In this paper, we generalize ACI, and the resulting operator is termed as generalized ACI (GACI). GACI adds to the efficacy of the ACI in the representation of a DM's unique and complex attitudinal character. It also generates a vast range of exponential ACI operators, such as harmonic ACI, ACI, quadratic ACI, to name a few. We further present induced GACI to consider additional information that may be associated with the arguments of aggregation. The special cases of the proposed operators are investigated. The usefulness of the proposed operators in modelling human decision behavior is shown through a case study.     

A Probabilistic and Decision Attitude Aggregation Operator for Intuitionistic Fuzzy Environment

In this paper, a new operator for aggregation of uncertain information under intuitionistic fuzzy environment is proposed. A novel approach is proposed for the selection of best alternative action in the face of the imprecise probabilities and the complex attitudinal character of the decision makers (DMs). This approach is distinguished with its capacity to accommodate the linguistic specification of probabilities as provided by human experts directly without the need to determine the fuzzy membership grades. The focus is to compute the net payoff for each alternative in the face of uncertain states of nature and DM's attitude. The proposed operator and the approach are illustrated through two real case studies.     

DISCRETE AND CONTINUOUS MODELS

Abstract

The growing influence of digital computing in systems modelling and simulation is leading to an increase in the use of discrete mathematical structures for describing models. While it is generally recognized that discrete methods and classical continuous methods both provide valuable tools for modelling, strong biases exist which depend on the modelling techniques that are traditional within specific disciplines. The choice of a modelling approach sometimes reflects the background of the model builder more strongly than it reflects the character of the problem to be solved. Since continuous methods have played the dominant role in scientific education, there are aspects of discrete modelling techniques and their relationship to continuous methods that are not widely recognized. The purpose of this article is to discuss some of these issues in order to dispel common criticisms of discrete techniques which are the result of unfamiliarity with discrete styles of mathematical thinking and a tendency to underestimate the degree of abstraction used in continuous approaches.     

7500吨浮吊齿轮箱故障诊断系统的研究

针对7500吨浮吊齿轮箱故障诊断问题,将离散小波变换和Tikhonov支持向量机结合建立了一个浮吊齿轮箱故障诊断系统。在输入层对振动信号进行离散小波变换,提取不同频带的能量参数作为故障特征向量,利用这些特征向量进行Tikhonov支持向量机的学习,训练后的Tikhonov支持向量机诊断浮吊齿轮箱故障。实验结果表明,离散小波Tikhonov支持向量机具有很强的故障识别性能和鲁棒性,诊断精度优于常规的BP网络方法。     

一种求解GP-决策树权值矢量算法及应用

提出一种基于遗传程序设计算法(GPA)求解决策树结点的权值矢量,并根据树结点的错误率与分割后的错误率减少量构造GP-决策树算法的方法。该方法不但可以求解出树结点的权值矢量,同时也确定了GP-决策树的结构。实验结果表明,应用GP-决策树算法能够正确完成对趋势预测模型的选择。     

一种概率序列核在说话人识别中的应用

以说话人识别中的背景模型为基础,根据模型中的各个高斯分量,构造出说话人特征空间,将长度不一样的语句映射成为空间中大小相同的向量,且经过相关矩阵进行规整后,采用线性支持向量机进行说话人识别。借鉴几种常见的特征规整方式,结合语句映射后的向量,提出四种不同的规整方法：均值/方差规整、权重规整、WLOG规整和球形规整,并与概率序列核进行比较研究。根据语音特征向量序列中相邻的特征向量的前后转移关系,结合提出的概率序列核,构造出转移概率序列核。实验在NIST2001库上进行,结果表明概率序列核模型识别性能接近经典的UBM-MAP模型,将这两类模型得分进行融合,可非常明显地提高识别性能,进一步融合转移概率序列核后,性能还可提高19.1%。     

Attitudinal Choquet integrals and applications in decision making

The compensation capabilities of Choquet integral are augmented to consider the complex attitudinal character of a decision maker. The resulting operator is termed as attitudinal Choquet integral (ACI). The proposed ACI is further extended as induced ACI. The special cases of ACI are investigated. The usefulness of ACI is shown through a case study.     

非确定选择平行模式的建模方法

非确定选择平行模式在实际业务流程中非常普遍,模式的执行行为根据案例属性和运行环境来动态选择,可以是平行、选择或者交替执行路由的模式,但是在许多情况下很难直接、有效地实现,很少有系统支持这种模式。基于同步管理器的流程建模方法从本质上将流程的管理任务和执行任务分离,明确了流程中稳定和可变的部分,这种分离有利于构建灵活的流程模型.不同的管理行为和模型结构重组可以构建非确定选择平行模式。实例分析表明基于同步管理器很容易建模非确定选择平行模式,而且语义表述更加精确,模型简单、有效,并具有较强的通用性。     

Analysis of the impact of attitudinal character on the multicriteria decision making with OWA operators

The first (or last) few alternatives need to be identified in some cases, such as recruiting (or dismissing) the first (or last) few candidates in a firm. In such cases, we analyze the impact of attitudinal character on the multicriteria decision making with ordered weighted averaging (OWA) operators. First, we propose a linear programming model to determine the best and worst ranking orders for each alternative to identify if it can be one of the potential first (or last) few alternatives. Second, we determine the attitudinal character's range for each alternative, in which it can be one of the potential alternatives. Furthermore, under the specified subintervals of attitudinal character, we identify which alternatives can be the potential ones by the analysis result obtained in the second step. According to the impact of attitudinal character analyzed by the proposed method, the set of alternatives can be reduced, when the range of attitudinal character is specified and the best and worst ranges of attitudinal character for each alternative can be identified in self‐evaluation. © 2012 Wiley Periodicals, Inc.     

A method for discrete stochastic MADM problems based on the ideal and nadir solutions

Many real life decision making problems can be modeled as discrete stochastic multi-attribute decision making (MADM) problems. A novel method for discrete stochastic MADM problems is developed based on the ideal and nadir solutions as in the classical TOPSIS method. In a stochastic MADM problem, the evaluations of the alternatives with respect to the different attributes are represented by discrete stochastic variables. According to stochastic dominance rules, the probability distributions of the ideal and nadir variates, both are discrete stochastic variables, are defined and determined for a set of discrete stochastic variables. A metric is proposed to measure the distance between two discrete stochastic variables. The ideal solution is a vector of ideal variates and the nadir solution is a vector of nadir variates for the multiple attributes. As in the classical TOPSIS method, the relative closeness of an alternative is determined by its distances from the ideal and nadir solutions. The rankings of the alternatives are determined using the relative closeness. Examples are presented to illustrate the effectiveness of the proposed method. Through the examples, several significant advantages of the proposed method over some existing methods are discussed.