基于粗糙集的多变量决策树构造方法

本文利用粗糙集理论中条件属性相对于决策属性的核，解决多变量检验中属性的选择问题．另外，定义了２个等价关系相对泛化的概念，并将它用于解决多变量检验的构造问题．通过一个例子，对本文提出的多变量决策树方法与著名的单变量决策树（ＩＤ３）方法进行了比较，结果表明前者比后者更简单．同时，对几种多变量决策树方法做了初步的对比分析．     

多决策树融合模型MDTF的研究

基于数据挖掘的入侵检测系统中存在着检测性能低和数据挖掘效率不高等问题.为了解决这些问题,提出了多决策树融合模型MDTF,也就是把海量数据集分成若干子数据集,在子数据集上进行挖掘形成不同的子决策树,然后用加权平均法将多棵子决策树对网络数据的检测结果进行融合形成最优判断.实验采用KDD99数据,实验结果表明,该方法可以得到较好的入侵检测性能,分布并行处理可以提高数据挖掘效率.     

基于遗传算法的多属性模糊决策树的优化

决策树是数据挖掘中的一种高效方法,但是当训练数据的属性很多时,构建的决策树的规模会随属性个数增加而指数级增长,进而会产生海量的规则。针对该问题,提出了一种基于遗传算法的优化方法。首先根据信息增益利用轮盘赌方法选取若干组属性,构建多棵决策树,然后利用遗传算法对多棵决策树进行组合,并最终形成规则集。最后给出了实验结果,证明了该方法的可行性和有效性。     

基于SVM多分类决策树的研究综述

支持向量机(Support Vector Machine,SVM)是一种基于统计学习理论的机器学习方法,由于其出色的学习性能,早已成为当前机器学习界的研究热点;而决策树是一种功能强大且相当受欢迎的分类和预测工具。本文重点介绍支持向量机与决策树结合解决多分类问题的算法,并对其进行评析和总结。     

构造性混合决策树

提出了一种构造性混合决策树学习方法CHDT。该方法用符号学习来进行定性分析,用神经学习进行后续的定量分析,在一定程度上模拟了人类的思维过程。CHDT采用了一种独特的构造性归纳机制,较好地解决了在缺乏领域知识指导的情况下进行构造性学习的问题,它通过采用FTART2网络和适宜于混合决策树的神经网络嵌入机制,获得了较强的泛化能力。实验结果表明,CHDT能构造出结构简洁、预测精度高的混合决策树。     

基于神经网络的模糊决策树改进算法

传统决策树通过对特征空间的递归划分寻找决策边界,给出特征空间的“硬”划分。但对于处理大数据和复杂模式问题时,这种精确决策边界降低了决策树的泛化能力。为了让决策树算法获得对不精确知识的自动获取,把模糊理论引进了决策树,并在建树过程中,引入神经网络作为决策树叶节点,提出了一种基于神经网络的模糊决策树改进算法。在神经网络模糊决策树中,分类器学习包含两个阶段：第一阶段采用不确定性降低的启发式算法对大数据进行划分,直到节点划分能力低于真实度阈值[ε]停止模糊决策树的增长;第二阶段对该模糊决策树叶节点利用神经网络做具有泛化能力的分类。实验结果表明,相较于传统的分类学习算法,该算法准确率高,对识别大数据和复杂模式的分类问题能够通过结构自适应确定决策树规模。     

基于多决策树算法的网络入侵检测

将一个大数据集分割成若干个子数据集，在每个子数据集上使用决策树算法进行挖掘，用投票的方式将多棵决策树的结果结合起来，形成全局的判断。将这种方法应用于网络入侵检测，试验表明，该方法不仅提高了数据挖掘算法对海量数据的处理能力，而且降低了误判率。     

改进的多关系决策树算法*

高效性和可扩展性是多关系数据挖掘中最重要的问题,而提高算法效率的主要瓶颈在于假设空间,且用户对分类的指导会在很大程度上帮助系统完成分类任务,减少系统独自摸索的时间。针对以上问题提出了改进的多关系决策树算法,即将虚拟连接元组传播技术和提出的背景属性传递技术应用到多关系决策树算法中。对改进的多关系决策树算法进行了理论证明,并且对多关系决策树算法和改进的多关系决策树算法进行比较实验。通过实验可以得出,当改进的多关系决策树在搜索数据项达到背景属性传递阈值时,改进的多关系决策树算法的效率相对很高且受属性个数增加(或     

基于改进决策树的电力通信设备状态预测算法研究

电力通信网络的快速增长,传统被动响应的事后运维模式迫切需要向分析预测的事前运维模式转型。论文在深入研究典型决策树理论的基础上,针对电力通信设备运行状态的关联特性,提出了一种改进的决策树学习算法,通过借鉴粗糙集理论对决策表属性的约简、求核与泛化过程,最终构造出一种多变量的决策树。通过算法仿真,该方法构建的决策树结构更为简化和合理,大大降低了计算量,提高了预测分析效率,为电力通信的运维提供了一种快速、简捷的通信设备状态预测方法,克服了经典决策树算法的不足,具有一定的实用性。     

基于主成分分析的多变量决策树构造方法

大多数决策树构造方法在每个节点上只检验单个属性,这种单变量决策树忽视了信息系统中广泛存在的属性间的关联作用,而且修剪时往往代价很大。针对以上两点,提出了一种基于主成分分薪的多变量决策树构造方法,提取信息系统中的若干主成分来构造决策树。实验结果表明,这是一种操作简单,效率很高的决策树生成方法。     

考虑风险偏好的区间直觉模糊软集型多属性决策方法

文章针对多方参与、评价指标有差异且存在模糊性的多属性方案优选问题,提出了一种考虑风险偏好的区间直觉模糊软集决策方法.对这类区间直觉模糊软集的多属性决策问题,提出了一个三阶段的决策方法,且按照决策者三种不同风险偏好,给出了相应的决策原则.在此基础上,文章选择了其中一种情形提出了具体的决策算法.最后,文章通过数值算例验证了该方法的可行性和有效性,同时采用综合得分值对该方法的结果进行了一致性的验证和讨论.该方法不仅可以很好地解决这类多属性方案优选问题,也能进一步推广到多属性方案的排序问题中.     

Design Of Robust Pole Assignment Based On Pareto‐Optimal Solutions

In this paper, a new design method for robust pole assignment based on Pareto‐optimal solutions for an uncertain plant is proposed. The proposed design method is defined as a two‐objective optimization problem in which optimization of the settling time and damping ratio is translated into a pole assignment problem. The uncertainties of the plant are represented as a polytope of polynomials, and the design cost is reduced by using the edge theorem. The genetic algorithm is applied to optimize this problem because of its multiple search property. In order to demonstrate the effectiveness of the proposed design method, we applied the proposed design method to a magnetic levitation system.     

检测识别跟踪分离的在线多样本视频目标跟踪

在视频目标跟踪中,经常出现“漂移”现象,且学习算法需要离线训练。为此,提出一种检测识别跟踪分离的在线多样本视频目标跟踪方法。利用多样本学习方法解决学习算法在更新过程中的内在不确定性,使用优于在线半监督的boosting方法解决“漂移”现象。实验结果表明,该方法鲁棒性较好,可以有效解决“漂移”现象,并能实时地完成在线跟踪。     

Commentary on “On solving a multiobjective fixed charge problem with imprecise fractional objectives”

Upmanyu and Saxena (Applied Soft Computing 40 (2016) 64–69) proposed a method for solving a multiobjective fixed charge problem having multiple fractional objective functions which are all of a fuzzy nature. The aim of this note is to aware the researchers that the method, proposed by Upmanyu and Saxena, is not valid and hence, to propose a method for solving this type of fixed charge problem is still an open challenging research problem.     

Optimal Load Balancing in a Multiple Processor System with Many Job Classes

A loosely coupled multiprocessor system contains multiple processors which have their own local memories. To balance the load among multiple processors is of fundamental importance in enhancing the performance of such a multiple processor system. Probabilistic load balancing in a heterogeneous multiple processor system with many job classes is considered in this study. The load balancing scheme is formulated as a nonlinear programming problem with linear constraints. An optimal probabilistic load balancing algorithm is proposed to solve this nonlinear programming problem. The proposed load balancing method is proven globally optimum in the sense that it results in a minimum overall average job response time on a probabilistic basis.     

A combined approach to nonlinear model predictive control of fast systems

A new solution strategy, a combination of the multiple shooting and collocation method, is proposed for nonlinear model predictive control (NMPC) of fast systems. The multiple shooting method is used for discretizing the dynamic model, through which the optimal control problem is transformed to a nonlinear program (NLP) problem. To solve this NLP problem the values of state variables and their gradients at the end of each shooting need to be computed. We use collocation on finite elements (CFE) to carry out this task. Due to its higher numerical accuracy the computation efficiency can be enhanced considerably, in comparison to an ordinary differential equation solver commonly used in the existing multiple shooting approach for integrating the ODEs and the chain-rule for the gradient computation. Therefore, the NMPC algorithm proposed can be applied to the control of fast systems. The performance of the proposed approach is demonstrated with three optimal control problems.     

Multiple robust H∞ controller design using the nonsmooth optimization method

This paper proposes a systematic technique to design multiple robust H_∞ controllers. The proposed technique achieves a desired robust performance objective, which is impossible to achieve with a single robust controller, by dividing the uncertainty set into several subsets and by designing a robust controller to each subset. To achieve this goal with a small number of divisions of the uncertainty set, an optimization problem is formulated. Since the cost function of this optimization problem is not a smooth function, a numerical nonsmooth optimization algorithm is proposed to solve this problem. This method avoids the use of Lyapunov variables, and therefore it leads to a moderate size optimization problem. A numerical example shows that the proposed multiple robust control method can improve the closed‐loop performance when a single robust controller cannot achieve satisfactory performance. Copyright © 2009 John Wiley & Sons, Ltd.     

安全芯片中密码算法的多IP核集成方法

基于目前SoC安全芯片多IP核集成的现实需要,在分析现有方法的基础上,设计并具体实现了一种多IP核集成方法。该方法通过IP桥,实现了不同IP核与同一双端口存储区的动态重构,有效整合了芯片内部资源,提高了芯片的整体性能。     

A new and efficient ant-based heuristic method for solving the traveling salesman problem

Abstract: In this paper, we present an efficient metaheuristic approach for solving the problem of the traveling salesman. We introduce the multiple ant clans concept from parallel genetic algorithms to search solution space using different islands to avoid local minima in order to obtain a global minimum for solving the traveling salesman problem. Our simulation results indicate that the proposed novel traveling salesman problem method (called the ACOMAC algorithm) performs better than a promising approach named the ant colony system. This investigation is concerned with a real life logistics system design which optimizes the performance of a logistics system subject to a required service level in the vehicle routing problem. In this work, we also concentrate on developing a vehicle routing model by improving the ant colony system and using the multiple ant clans concept. The simulation results reveal that the proposed method is very effective and potentially useful in solving vehicle routing problems.     

基于遗传算法的多旅行商问题研究

针对所有旅行商路径总和最小为优化标准的多旅行商一类问题,用遗传算法优化,并提出了矩阵解码方法。对距离非对称的多旅行商问题的实例进行了仿真,并对不同交叉算子性能进行了比较。结果表明,该算法是有效的,适用于距离对称和非对称的多旅行商问题求解。