Multi-Class Support Vector Machine Classifier Based on Jeffries-Matusita Distance and Directed Acyclic Graph

Based on the framework of support vector machines （SVM） using one-against-one （OAO） strategy, a new multi-class kernel method based on directed aeyclie graph （DAG） and probabilistic distance is proposed to raise the multi-class classification accuracies. The topology structure of DAG is constructed by rearranging the nodes＇ sequence in the graph. DAG is equivalent to guided operating SVM on a list, and the classification performance depends on the nodes＇ sequence in the graph. Jeffries-Matusita distance （JMD） is introduced to estimate the separability of each class, and the implementation list is initialized with all classes organized according to certain sequence in the list. To testify the effectiveness of the proposed method, numerical analysis is conducted on UCI data and hyperspectral data. Meanwhile, comparative studies using standard OAO and DAG classification methods are also conducted and the results illustrate better performance and higher accuracy of the orooosed JMD-DAG method.

Multi-Fault Diagnosis for Autonomous Underwater Vehicle Based on Fuzzy Weighted Support Vector Domain Description

This paper addresses the multi-fault diagnosis problem of thrusters and sensors for autonomous underwater vehicles (AUVs). Traditional support vector domain description (SVDD) has low classification accuracy in the process of AUV multi-fault pattern classification because of the effect of sample sparse density and the uneven distribution of samples, and so on. Thus, a fuzzy weighted support vector domain description (FWSVDD) method based on positive and negative class samples is proposed. In this method, the negative class sample is introduced during classifier training, and the local density and the class weight are introduced for each sample. To improve the multi-fault pattern classifier training speed and fault diagnosis accuracy of FWSVDD, a multi-fault mode classification method based on a hierarchical strategy is proposed. This method adds fault contain detection surface for each thruster and sensor to isolate fault components during fault diagnosis. By considering the problem of pattern classification for a fuzzy sample, which may be located in the overlapping area of hyper-spheres or may not belong to any hyper-sphere in the process of multi-fault classification based on FWSVDD, a relative distance judgment method is given. The effectiveness of the proposed multi-fault diagnosis approach is demonstrated through water tank experiments with an experimental AUV prototype.

Fault Diagnosis for Aero-engine Applying a New Multi-class Support Vector Algorithm

Hierarchical Support Vector Machine （H-SVM） is faster in training and classification than other usual multi-class SVMs such as ＂1-V-R＂and ＂1-V-1＂. In this paper, a new multi-class fault diagnosis algorithm based on H-SVM is proposed and applied to aero-engine. Before SVM training, the training data are first clustered according to their class-center Euclid distances in some feature spaces. The samples which have close distances are divided into the same sub-classes for training, and this makes the H-SVM have reasonable hierarchical construction and good generalization performance. Instead of the common C-SVM, the v-SVM is selected as the binary classifier, in which the parameter v varies only from 0 to 1 and can be determined more easily. The simulation results show that the designed H-SVMs can fast diagnose the multi-class single faults and combination faults for the gas path components of an aero-engine. The fault classifiers have good diagnosis accuracy and can keep robust even when the measurement inputs are disturbed by noises.

Data fusion for fault diagnosis using multi-class Support Vector Machines

Multi-source multi-class classification methods based on multi-class Support Vector Machines and data fusion strategies are proposed in this paper. The centralized and distributed fusion schemes are applied to combine information from several data sources. In the centralized scheme, all information from several data sources is centralized to construct an input space. Then a multi-class Support Vector Machine classifier is trained. In the distributed schemes, the individual data sources are proc-essed separately and modelled by using the multi-class Support Vector Machine. Then new data fusion strategies are proposed to combine the information from the individual multi-class Support Vector Machine models. Our proposed fusion strategies take into account that an Support Vector Machine (SVM) classifier achieves classification by finding the optimal classification hyperplane with maximal margin. The proposed methods are applied for fault diagnosis of a diesel engine. The experimental results showed that almost all the proposed approaches can largely improve the diagnostic accuracy. The robustness of diagnosis is also improved because of the implementation of data fusion strategies. The proposed methods can also be applied in other fields.

Progressive transductive learning pattern classification via single sphere

In many machine learning problems, a large amount of data is available but only a few of them can be labeled easily. This provides a research branch to effectively combine unlabeled and labeled data to infer the labels of unlabeled ones, that is, to develop transductive learning. In this article, based on Pattern classification via single sphere （SSPC）, which seeks a hypersphere to separate data with the maximum separation ratio, a progressive transductive pattern classification method via single sphere （PTSSPC） is proposed to construct the classifier using both the labeled and unlabeled data. PTSSPC utilize the additional information of the unlabeled samples and obtain better classification performance than SSPC when insufficient labeled data information is available. Experiment results show the algorithm can yields better performance.

Multi-class classification method for strip steel surface defects based on support vector machine with adjustable hyper-sphere

Focusing on strip steel surface defects classification, a novel support vector machine with adjustable hyper-sphere （AHSVM） is formulated. Meanwhile, a new multi-class classification method is proposed. Originated from support vector data description, AHSVM adopts hyper-sphere to solve classification problem. AHSVM can obey two principles： the margin maximization and inner-class dispersion minimization. Moreover, the hyper-sphere of AHSVM is adjustable, which makes the final classification hyper-sphere optimal for training dataset. On the other hand, AHSVM is combined with binary tree to solve multi-class classification for steel surface defects. A scheme of samples pruning in mapped feature space is provided, which can reduce the number of training samples under the premise of classification accuracy, resulting in the improvements of classification speed. Finally, some testing experiments are done for eight types of strip steel surface defects. Experimental results show that multi-class AHSVM classifier exhibits satisfactory results in classification accuracy and efficiency.

基于多域特征融合的旋翼无人机分类识别

为提高雷达旋翼无人机的识别效果,本文提出一种基于多域特征融合的旋翼无人机分类方法。首先利用K波段连续波(Continuous Wave,CW)雷达观测多旋翼无人机,对采集到的雷达回波信号进行信号处理依次得到时频图、节奏速度图(Cadence⁃Velocity Diagram,CVD)和节奏频谱图(Cadence Frequency Spectrum,CFS),然后将时频图和CVD图分别输入SqueezeNet网络,CFS数据输入一维卷积神经网络(1⁃D⁃CNN)提取回波信号在时频域、节奏速度域和节奏频率域的特征,最后将特征融合输入支持向量机(Support Vector Machine,SVM)进行分类。实测雷达数据处理的结果表明基于多域特征融合的旋翼无人机分类识别方法对三类旋翼无人机的分类准确率达到99.14%。

Multi-class Classification Methods of Enhanced LS-TWSVM for Strip Steel Surface Defects

Considering strip steel surface defect samples, a multi-class classification method was proposed based on enhanced least squares twin support vector machines （ELS-TWSVMs） and binary tree. Firstly, pruning region samples center method with adjustable pruning scale was used to prune data samples. This method could reduce classifierr s training time and testing time. Secondly, ELS-TWSVM was proposed to classify the data samples. By introducing error variable contribution parameter and weight parameter, ELS-TWSVM could restrain the impact of noise sam- ples and have better classification accuracy. Finally, multi-class classification algorithms of ELS-TWSVM were pro- posed by combining ELS-TWSVM and complete binary tree. Some experiments were made on two-dimensional data- sets and strip steel surface defect datasets. The experiments showed that the multi-class classification methods of ELS-TWSVM had higher classification speed and accuracy for the datasets with large-scale, unbalanced and noise samples.

Multi-class classification method for steel surface defects with feature noise

Defect classification is the key task of a steel surface defect detection system.The current defect classification algorithms have not taken the feature noise into consideration.In order to reduce the adverse impact of feature noise,an anti-noise multi-class classification method was proposed for steel surface defects.On the one hand,a novel anti-noise support vector hyper-spheres(ASVHs)classifier was formulated.For N types of defects,the ASVHs classifier built N hyper-spheres.These hyper-spheres were insensitive to feature and label noise.On the other hand,in order to reduce the costs of online time and storage space,the defect samples were pruned by support vector data description with parameter iteration adjustment strategy.In the end,the ASVHs classifier was built with sparse defect samples set and auxiliary information.Experimental results show that the novel multi-class classification method has high efficiency and accuracy for corrupted defect samples in steel surface.

基于支持向量机元分类器的体育视频分类

为弥补特征提取中的语义缺陷,提出了一种利用领域知识规则填补特征与高级语义之间鸿沟的思想,从体育视频中对语义对象进行有效的特征提取,并采用支持向量机元分类器和组合策略对体育视频进行分类的方法.实验表明,该分类方法对大部分体育视频都具有很好的分类效果,平均准确率可达92.23%,优于其他提取特征无语义关联的分类方法.

风电功率爬坡气象场景分类模型及阈值整定研究

为了判定和预报引发风电功率爬坡事件的风速突变的强对流气象类型,考虑风电场实际运行状态、电力系统运行方式以及区域电网的热备用启动速度和承受能力确定风电功率爬坡定义及其爬坡气象场景判定标准。在此定义上,引入支持向量标记法构造了风电功率爬坡场景分类的极小值和极大值的初始化模型,通过合适的显著性参数因子及预分类结果,建立风电功率爬坡场景分类模型。进而根据气象学物理意义分类出典型爬坡气象类型和相关特征因子阈值范围。实例分析表明,风电功率爬坡气象场景分类模型和确定的分类特征因子阈值对预报判别出目标区域风电功率爬坡气象类型有较好的指导作用。

基于信号特征空间的TDCS干扰分类识别

针对变换域通信系统中干扰信号的分类识别问题,提出了一种基于信号特征空间的支持向量机(signal feature space-support vector machine,SF-SVM)干扰分类算法。首先,基于干扰信号模型和信号空间理论对干扰信号进行特征提取,并建立信号特征空间,进而针对二分类和多分类问题提出了SF-SVM分类算法,设计了干扰信号的多分类识别器。仿真结果表明,与干扰信号的传统分类算法相比,SF-SVM不仅提高了分类精度,而且缩短了训练时间;设计的多分类识别器在信噪比达到8dB时,对6种干扰信号识别性能及对变换域通信系统性能都有所提升。

基于SVDD的层次纠错输出编码研究

纠错输出编码能有效地将多类问题分解为一系列二类子问题进行求解,已受到众多机器学习研究者的关注。如何构建基于数据的编码矩阵是编码方法确定的关键。针对此问题,基于Fisher原理,提出一种基于支持向量数据描述(support vector domain description,SVDD)的层次纠错输出编码构造方法(hierarchical error-correcting output codes,HECOC)。该方法首先采用SVDD计算各类别的可分程度,从而得到由不同子类构成的二叉树;然后分别对二叉树的各层结点进行编码并最终形成层次输出编码。在仿真实验中,对不同子类类群划分构成的基分类器的可分性进行了对比,结果表明,该编码方法能在保证分类精度的同时,提高基分类器之间的差异性和纠错输出编码的容错能力。

领域适应核支持向量机

领域适应学习是一种新颖的解决先验信息缺少的模式分类问题的有效方法,最大化地缩小领域间样本分布差是领域适应学习成功的关键因素之一,而仅考虑领域间分布均值差最小化,使得在具体领域适应学习问题上存在一定的局限性.对此,在某个再生核Hilbert空间,在充分考虑领域间分布的均值差和散度差最小化的基础上,基于结构风险最小化模型,提出一种领域适应核支持向量学习机(Kernel support vector machine for domain adaptation,DAKSVM)及其最小平方范式,人造和实际数据集实验结果显示,所提方法具有优化或可比较的模式分类性能。

基于EEMD和进化支持向量机的齿轮混合智能诊断方法研究

针对齿轮早期故障特征不明显,提出了一种基于总体平均经验模态分解(ensemble empirical mode decomposition,EEMD)和进化支持向量机相结合的齿轮故障智能诊断方法。利用EEMD能对齿轮振动信号进行自适应的分解成若干本征模式分量(intrinsic mode function,IMFs),并能有效抑制经典经验模式分解可能出现的模式混叠现象。以所得的IMF分量中提取出来的能量特征为输入建立进化支持向量机,判断齿轮的故障状态。结果表明:建立的混合智能诊断方法的分类正确率最高,能有效诊断齿轮早期故障。

融合时空上下文的手绘笔画图文分类

在笔式用户界面中,对手绘图形和手写文字的识别通常采用不同的识别算法,因此通过笔画分类将混杂的笔画集自动分离是手绘草图识别中的一个重要研究课题。该文提出一种融合时空上下文的手绘笔画联合分类方法,采用支持向量随机场对时空关联的笔画集进行联合建模,不仅利用笔画自身的特征进行判别分类,还以时空邻域和笔画对特征同时融合了笔画间的时间上下文和空间上下文信息,通过模型环状置信传播(LBP)推断,最终求得最大后验边缘概率准则下的联合分类结果。实验结果表明,该文方法的分类准确率优于基于SVM的单笔画分类方法和基于马尔科夫随机场(MRF)的空间上下文联合分类方法,分类速度能基本满足交互实时性的要求,验证了利用随机场模型融合时空上下文进行笔画分类的可行性和有效性。

基于GSM_SVDD的模拟电路故障诊断方法

在基于支持向量数据描述(support vector domain description,SVDD)的模拟电路故障诊断中,故障样本易陷入多个球体的交叉区域产生误诊。为了改进标准SVDD松弛的球体描述边界以提高故障诊断性能,提出一种基于图谱空间映射SVDD(graph spectrum mapping SVDD,GSM_SVDD)的模拟电路故障诊断新方法。采用高斯核函数构造Laplace矩阵,然后进行特征值分解,由特征值对应的Laplace特征向量描述SVDD球体的边界,最后采用SVDD的最小相对距离法则诊断故障样本。实验结果表明,通过Laplace谱映射改变原始特征样本的空间分布,GSM-SVDD方法能有效提高模拟电路的故障诊断性能。

非平衡数据集的支持向量域分类预测模型研究

基于非平衡数据集的支持向量域分类模型,提出了一种银行客户个人信用预测方法。首先分析了信用预测的主要方法及其不足,然后研究了支持向量域分类模型及其参数的非负二次规划乘性更新算法,进而提出基于支持向量域分类模型的银行客户个人信用预测方法,最后使用人工数据和实际数据对提出方法与支持向量机预测方法进行对比实验。实验结果表明对于银行客户个人信用预测的非平衡数据分析问题,基于支持向量域模型的分类预测方法更有效。

ECOC多分类器实现的最小封闭球模型

基于纠错输出码(error-correcting output codes,ECOC)的多分类器实现旨在通过构造多个二分类器,根据各个二分类器的输出对测试样本进行分类决策,标准的做法是采用最短海明距离判别.首先对传统二进制ECOC的多分类模型进行了几何刻画,给出了ECOC多分类器的最小封闭球几何描述模型,然后把这种思想推广到实数编码的实现,并采用支持向量域描述(support vector domain description,SVDD)在实数向量空间中寻找各个类别的最小封闭球.进一步根据最小封闭球的几何模型,探讨了给出后验概率估计的ECOC多分类器实现策略.最后采用支持向量机作为ECOC的二类分类器,在UCI数据集上进行了实验分析.实验结果表明:对于长度较短的ECOC编码,所提出的计算模型在分类精度上相比传统的方法性能明显改善.

多运动形式下的行人三维定位方法研究

针对行人复杂多变的运动形式给室内定位带来较大偏差的问题,提出了一种基于加速度时域特征的行人运动分类方法,并利用分类结果进行室内行人三维定位。利用垂直加速度的变化规律将加速度信号划分为连续的单步信号,计算单步周期内加速度信号的时域特征,基于BP神经网络和支持向量机设计一种二分树结构的分类器。经大量人员运动实验验证,该分类器对走、跑,上下楼梯3类运动形式的分类准确率接近100%,上、下楼梯的分类准确率为95%;在行人运动形式确定的情况下,利用不同的步长模型和航向信息进行室内三维定位,定位误差为1.5 m。