An OWA-TOPSIS method for multiple criteria decision analysis

A hybrid approach integrating OWA (Ordered Weighted Averaging) aggregation into TOPSIS (technique for order performance by similarity to ideal solution) is proposed to tackle multiple criteria decision analysis (MCDA) problems. First, the setting of extreme points (ideal and anti-ideal points) in TOPSIS is redefined and extended for handling the multiple extreme points situation where a decision maker (DM) or multiple DMs can provide more than one pair of extreme points. Next, three different aggregation schemes are designed to integrate OWA into the TOPSIS analysis procedure. A numerical example is provided to demonstrate the proposed approach and the results are compared for different aggregation settings and confirm the robustness of rankings from different scenarios.     

Decision tree induction based on minority entropy for the class imbalance problem

Most well-known classifiers can predict a balanced data set efficiently, but they misclassify an imbalanced data set. To overcome this problem, this research proposes a new impurity measure called minority entropy, which uses information from the minority class. It applies a local range of minority class instances on a selected numeric attribute with Shannon’s entropy. This range defines a subset of instances concentrating on the minority class to be constructed by decision tree induction. A decision tree algorithm using minority entropy shows improvement compared with the geometric mean and F-measure over C4.5, the distinct class-based splitting measure, asymmetric entropy, a top–down decision tree and Hellinger distance decision tree on 24 imbalanced data sets from the UCI repository.     

基于模糊多准则决策的谈判模型

针对目前谈判支持系统中谈判模型存在的难以准确获知谈判人的偏好,以及利用偏好构造其效用函数的问题,应用模糊数学的方法,将谈判人的偏好和对目标值的评价用梯形模糊数表示,由此提出一种基于模糊多准则决策方法求解谈判问题的模型。通过示例分析计算,表明该模型能帮助谈判各方达成一个满意解。     

Microcomputer based multiple criteria decision support system for strategic planning

There are two major approaches currently used for developing Decision support Systems (DSS) for strategic planning, especially in the objective formulation stage. Several mathematical models have been developed to abstract the decision situation. However, they do not take into account either behavioral aspects of decision making or the presence of multiple and conflicting objectives. A second approach is to consider the several qualitative factors that go into decision making; such considerations are normally situation-dependent and hence it is difficult to provide a system for general managerial situations.The Multiple Criteria Decision Making (MCDM) approach combines the advantages of both the approaches and, therefore, is an excellent alternative for designing DSS. This paper develops an MCDM approach to strategic planning. The model is applied to such a problem in a simulated environment and the problem is solved interactively. Our experience shows that the proposed methodology is a viable approach for solving practical decision problems in strategic planning.     

基于相关性的类偏好敏感决策树算法

针对决策者在面对几个分类结果时会有选择其中某一个结果的倾向性这一事实,提出了一种基于相关性的类偏好敏感决策树分类算法（CPSDT）。该算法引入了类偏好度、偏好代价矩阵等概念。为弥补在传统决策树构造过程中,选择分裂属性时未考虑非类属性之间相关性的不足,该算法在进行学习之前先采用基于相关性的特征预筛选排除属性冗余并重新构造了基于相关性的属性选择因子。经实验证明,该算法能够有效减小决策树规模,且能够在实现对偏好类的高精度预测的同时保证决策树拥有较好的整体精度。     

基于模糊分裂的概念自适应增量决策树分类算法

针对数据流环境下传统分类挖掘算法的不足,引入了改进的滑动窗口技术和模糊技术,通过在滑动窗口中设置分类效用因子的方法提高了窗口的利用率,有效改善了由于概念漂移所带来的分类器过时的问题,在连续属性分裂过程中加入了模糊技术,解决了连续属性字段平滑离散化的问题。理论分析和实例表明了改进后的算法具有较低的运行环境要求和较高的分类准确率。     

Cost-sensitive collaborative representation based classification via probability estimation with addressing the class imbalance

Multimedia Tools and Applications - Collaborative representation has been successfully used in pattern recognition and machine learning. However, most existing collaborative representation...     

A method for stochastic multiple criteria decision making based on dominance degrees

This paper proposes a method for solving the stochastic multiple criteria decision making (SMCDM) problem, where consequences of alternatives with respect to criteria are represented by random variables with probability distributions. Firstly, definitions and related analysis of dominance degree of one probability distribution over another are given. Then, by calculating the dominance degrees, the dominance degree matrix of alternative pairwise comparisons with respect to each criterion is built. Further, using PROMETHEE II method, an overall dominance degree matrix of alternative pairwise comparisons is constructed, and a net flow of each alternative is calculated. Based on the obtained net flows, a ranking of alternatives is determined. Finally, numerical examples for the three cases are given to illustrate the use of the proposed method.     

An adaptive group decision making framework: Individual and local world opinion based opinion dynamics

Opinion dynamics (OD) models, which simulate individuals’ opinion evolution process on social network to analyze the final state of opinion distribution in a group, usually differ from each other due to the differences in social network evolution rules and opinion evolution rules. However, most existing social network evolution rules and opinion evolution rules usually cannot characterize the comprehensive influence of key factors such as neighbors and opinion differences in social relationships. To fully consider the properties of social network evolution and improve the efficiency of consensus reaching process in group decision making, this paper introduces the concept of local world opinion derived from individuals’ common friends, and then proposes an individual and local world opinion-based OD model. In the proposed model, social network evolution is jointly determined by the distance between individual opinions and network structure similarity. The pair of individuals with the largest consensus improvement space are then suggested to adjust their opinions by using an adaptive individual opinion adjustment mechanism. Finally, detailed simulation results are provided to demonstrate the convergence of the proposed model and analyze different parameters’ effects on the stabilized time steps and the number of stable state opinion clusters.     

A method based on stochastic dominance degrees for stochastic multiple criteria decision making

In this paper, a novel method based on the stochastic dominance degree (SDD) is proposed to solve a discrete stochastic multiple criteria decision-making (MCDM) problem. Firstly, a concept of stochastic dominance degree is introduced to describe the degree that one alternative dominates another when the SD relation for each pair of alternatives is determined, and a computation formula of the SDD is given. Then, by calculating SDDs, the SDD matrix on pairwise comparisons of alternatives with respect to each criterion is built. Furthermore, the SDD matrices with respect to all the criteria are aggregated into an overall SDD matrix using the simple additive weighting method. Based on the overall SDD matrix, an approach based on the idea of the PROMETHEE-II is developed to obtain the ranking result of alternatives. Finally, two numerical examples are used to illustrate the applicability and effectiveness of the proposed method.     

基于MapReduce的Bagging决策树优化算法

针对经典C4.5决策树算法存在过度拟合和伸缩性差的问题,提出了一种基于Bagging的决策树改进算法,并基于MapReduce模型对改进算法进行了并行化。首先,基于Bagging技术对C4.5算法进行了改进,通过有放回采样得到多个与初始训练集大小相等的新训练集,并在每个训练集上进行训练,得到多个分类器,再根据多数投票规则集成训练结果得到最终的分类器;然后,基于MapReduce模型对改进算法进行了并行化,能够并行化处理训练集、并行选择最佳分割属性和最佳分割点,以及并行生成子节点,实现了基于MapReduce Job工作流的并行决策树改进算法,提高了对大数据集的分析能力。实验结果表明,并行Bagging决策树改进算法具有较高的准确度与敏感度,以及较好的伸缩性和加速比。     

基于代价敏感混合分裂策略的多决策树算法

煤矿瓦斯预警可视为是否安全的分类问题,数据呈现不平衡分布特点。为此,提出一种混合策略属性选择多决策树分类算法:算法融合代价敏感因子,结合C4.5和CART属性选择方法作为分裂指标,并采用了基于不同根节点信息的多决策树建树方法。首先采用11个非平衡数据集进行算法有效性验证,实验结果表明,该方法可以有效针对不平衡数据进行分类,保证高准确率的前提下,有效提高了少数类预测准确性;进而将该算法用于煤矿瓦斯数据预测,结果表明,所提出方法可以有效提高煤矿瓦斯数据的总体预测性能。     

基于赤池信息准则的分类回归决策树剪枝算法

决策树剪枝可以提高决策树的分类准确度。代价复杂度剪枝（ CCP）等常用的剪枝算法,都以降低决策树的误判率作为剪枝依据。引入赤池信息准则（ AIC）评价决策树的优良性,并提出了基于AIC的决策树剪枝算法,将分类正确概率和复杂度的综合评价作为剪枝依据。通过实例分析,基于AIC的剪枝算法能够得到高分类准确度的决策树,并没有出现过拟合或剪枝不充足等问题。     

多属性决策中赋权方法的自适应选择研究

为解决多属性决策中众多赋权方法难以选择的问题,对主客观权值向量冲突因子与评估结果精度的相关性进行了分析,进而提出了一种基于冲突阈值的自适应选择算法。通过实验仿真,结果表明：该算法消除了主、客观权值向量冲突对评估结果的影响,提高了计算精度,实现了基于冲突阈值的自适应选择。而且,当冲突因子小于阈值时,自适应选择算法将选择算法时间复杂度低于乘法合成原理的综合赋权法。当冲突因子大于阈值时,选择计算结果准确度高的基于博弈论的综合赋权法。     

网络流量识别的自适应分级滑动窗决策树算法

针对网络流量存在概念漂移、不同应用类型数据流偏态分布等特性, 提出了基于Hoeffding决策树的自适应分级滑动窗决策树的网络流量识别算法。该算法根据节点信息增益率检测概念漂移、动态调整概念漂移检测窗口及不同类型训练样本集窗口, 实现对不同速率概念漂移的自适应分类和决策树更新。实验结果显示新算法对劣势频繁漂移的应用类型的识别准确率与batch C4. 5算法接近, 比CVFDT算法提高约20%, 可以获得更加均衡的不同应用类型分类准确度。     

基于云发生算法的犹豫语言多准则决策方法

犹豫语言集比传统语言变量能更好地表示不确定决策信息,但因其具有非均匀和离散的特点,使用时难以找到对应的解析词运算规则和比较方法。对此,利用云模型的云滴生成算法,通过数值模拟方法解决犹豫语言信息的运算与比较问题,提出了相应的多准则决策方法,并通过算例分析了其有效性。     

考虑准则依赖的风险型多准则决策新方法

为有效应对风险型多准则决策过程中准则依赖的问题,针对现有文献中完全理性和有限理性决策两种视角下准则独立假设的不足,应用DEMATEL方法分析决策准则之间的相互影响关系,得到准则依赖情境下的综合重要性权重,进而以备选方案状态均值为参照点,从前景理论出发,求解风险状态下备选方案的综合前景价值,作为决策依据。案例应用结果表明,所提方法在实践中是切实可行的。     

A new multiple criteria decision making method based on intuitionistic fuzzy information

The risk attitude of a decision maker is considered in the decision process. Inspired by mean-variance type utility functions in the financial risk management, a new class of decision functions are defined based on the weighted score function and the weighted accuracy function in the intuitionistic fuzzy setting. By choosing a suitable parameter value, the decision maker’s risk attitude can be flexibly reflected by our decision function. The new method can be applied for both the exactly known and partly known criteria weight situations. For the latter case, it is only necessary to solve one linear programming problem. The developed models and algorithms are then extended to multiple criteria decision making problems with the interval-valued intuitionistic fuzzy information. Numerical examples are provided to illustrate the practicality, flexibility and efficiency of our new algorithms.     

基于相关系数的决策树优化算法

通过分析ID3算法的基本原理及其多值偏向问题,提出了一种基于相关系数的决策树优化算法。首先通过引进相关系数对ID3算法进行改进,从而克服其多值偏向问题,然后运用数学中泰勒公式和麦克劳林公式的性质,对信息增益公式进行近似简化。通过具体数据的实例验证,说明优化后的ID3算法能够解决多值偏向问题。标准数据集UCI上的实验结果表明,在构建决策树的过程中,既提高了平均分类准确率,又降低了构建决策树的复杂度,从而还缩短了决策树的生成时间,当数据集中的样本数较大时,优化后的ID3算法的效率得到了明显的提高。     

一类基于多目标规划的线性鉴别准则

两类Fisher鉴别准则、大间距线性投影准则以及最大散度差鉴别准则都是直接用于模式分类的两类线性鉴别准则,它们的共同点是将“投影后数据的可分性达到最大的方向”作为最优投影方向。区别在于它们对数据可分性的定义有所不同。过去的研究成果表明,大间距线性投影分类器与支持向量机之间、大间距线性投影准则与最大散度差鉴别准则之间以及最大散度差鉴别准则与两类Fisher鉴别准则之间,均存在着这样或那样的联系。论文试图在以往研究成果的基础上进一步理清这些两类线性鉴别准则之间的内在关系,并建立一个统一的理论框架从而将基于投影后数据可分性的这些两类线性鉴别准则都纳入其中。