一种基于遗传算法的SVM决策 树多分类策略研究

提出了一种基于遗传算法(GA)的SVM最优决策树生成算法,并将其应用于解决SVM多分类问题.首先以最大分类间隔为准则,利用遗传算法对传统的SVM决策树进行优化,生成最优(或近优)决策二叉树;然后在各个决策节点,利用传统的SVM算法进行二值分类,最终实现SVM的多值分类.理论分析及实验结果表明,新方法比传统的DT-SVM、DAG-SVM方法有更高的分类精度,比经典的1-a-1、1-a-r有更高的训练和分类效率.     

一种基于二值粒子群优化和支持向量机的目标检测算法

针对复杂场景下目标检测和目标检测中特征选择问题,该文将二值粒子群优化算法(BPSO)用于特征选择,结合支持向量机(SVM)技术提出了一种新颖的基于BPSO-SVM特征选择的自动目标检测算法。该算法将目标检测转化为目标识别问题,采用wrapper特征选择模型,以SVM为分类器,通过样本训练分类器,根据分类结果,利用BPSO算法在特征空间中进行全局搜索,选择最优特征集进行分类。基于BPSO-SVM的特征选择方法降低了特征维数,显著提高了分类器性能。实验结果表明,该文算法不仅有效提高了复杂场景下目标姿态、尺度、光照变化和局部被遮挡时的检测准确率,还大大缩短了检测时间。     

非均衡数据目标识别中SVM模型多参数优化选择方法

提出了非均衡数据目标识别中SVM模型多参数优化选择方法.首先从理论上分析了SVM模型多参数选择的内涵和必要性,针对非均衡数据的分类识别,基于F测度提出了能全面反映识别性能的多参数优化选择准则.在多参数选择过程中,利用遗传算法进行模型多参数并行优化选择.提出的方法能够寻找模型多参数的全局最优解,避免陷入梯度法常出现的局部最优解情况,同时能够克服传统方法中根据经验选择SVM单参数模型时计算量太大的不足.采用国际通用的标准数据集和雷达目标HRRP数据集进行了仿真实验,实验结果表明,该方法能够得到模型多参数的全局最优值,由此确定的SVM模型分类器性能有较大提高.     

支持向量机方法在文本分类中的改进

提出了一种应用于文本分类的KNN和SVM相结合的算法,将SVM近似看成每类只有一个代表点的1NN分类器,对于待识别样本,如果其离支持向量机的最优分界面较远,则用SVM分类;如果其离分界面较近,采用KNN对测试样本分类,将每个支持向量作为代表点,计算待识别样本和每个支持向量的距离对其作出判断.该算法综合了KNN和SVM在分类问题中的优势,既有效地降低了分类候选的数目,又提高了文本分类的精度.最后用实验验证了该算法的有效性.     

基于SVM算法的林果害虫图像识别

支持向量机已经成功应用在许多领域如规则提取、分类和评价。针对阿克苏林果害虫分类种类多的特点,采用了一种简单sift特征提取,结合svm算法分类识别林果主要害虫的方法。首先对样本数据进行SVM分类器训练,从而快速获取原始样本数据的最佳SVM分类器参数,并实现分类。利用神经网络算法进行分类实验,比较分析传统SVM算法、神经网络算法在分类识别方面的差异。实验结果表明,SVM算法能够快速有效地识别害虫种类,并获得相对较高的分类精度,但由于测试图像背景不同且提取的特征单一识别的准确率有待提高。     

基于可变特征空间SVM的互联网流量分类

支持向量机（support vector machine,SVM）是一类具有良好泛化能力的机器学习算法,适合应用于互联网动态环境下的流量分类问题。目前将SVM扩展到流量分类这样的多分类问题的方法主要有One-Against-All和One-Against-One方法。这些方法都基于单一的特征空间训练SVM两分类器,没有考虑到不同特征对不同流量类的不同区分能力,因此获得的分离超平面并不是最合理的。为此提出了可变特征空间的SVM集成方法,即为每个两分类 SVM 构建具有最优区分能力的独立特征空间,单独训练两分类 SVM,最后再利用One-Against-All和One-Against-One方法集成为多分类器。实验表明,与原来的单一特征空间的One-Against-All和One-Against-One集成方法相比,提出的方法能有效提高流量分类器分类精度和召回率,更易获得最优分离超平面。     

改进的AdaBoost.M2-SVM在低信噪比语音识别中的应用

提出了基于雁群启示的粒子群优化算法改进的AdaBoost.M2-SVM算法.首先训练多个支持向量机作为弱分类器,用AdaBoost.M2算法将多个弱分类器集成为最终的强分类器,实现多类分类;采用GeesePSO算法对AdaBoost.M2算法计算出的权值进行优化得到一组最优的权值,提高最终强分类器的提升能力.实验结果表明,在低信噪比语音识别中,与SVM相比,改进的AdaBoost.M2-SVM表现出更好的泛化能力,提高了识别准确率.     

Boosting优化决策树的带钢表面缺陷识别技术

基于图像信息的缺陷识别技术是带钢表面缺陷检测系统中的关健技术之一.通过采用单一的分类技术或者一步到位的创建分类器,对复杂带钢表面缺陷图像进行识别很难达到理想的效果.提出了用Boosting算法结合SLIQ决策树建立组合分类器来识别带钢表面缺陷的方法.Boosting算法通过适应性权重技术和带权重的投票方法,建立并组合多个功能互补的分类器,组合分类器通过优势互补的方法有效地提高单个分类器的性能;而SLIQ决策树算法的数据预排序和广度优先技术对大规模数据分类具有速度优势,适合于作为单个分类器的弱学习算法.对实际带钢表面缺陷数据集进行测试,Boosting优化SLIQ决策树的组合分类器对缺陷识别的准确率达到了90%以上.     

基于多特征提取和粒子群算法的图像分类

现有图像分类大都采用单一特征,不能利用多个特征之间性能互补优势,且将特征选择与分类器构造分割开来,影响图像分类的精度和分类器的泛化能力。针对以上问题提出一种基于混沌二进制粒子群算法(CBPSO)的特征选择和SVM参数同步优化方法,利用图像的综合特征,将特征选择和SVM分类器构造结合同步优化,仿真实验结果表明,该算法能同步找出最优的特征子集和合适的SVM参数,提高了图像分类精度和分类器泛化能力。     

数字调制识别SVM分类器参数优化及仿真

对于基于SVM数字信号调制识别分类器,参数选取过程中如何优化惩罚因子和径向基核函数参数问题,提出了一种改进算法。该算法将自适应惯性权重粒子群算法和k折交叉验证法结合,利用交叉验证法计算粒子适应度值,通过粒子群算法实现最优参数值搜索,最终得到分类器惩罚因子和径向基核函数参数最优值。仿真结果表明,该算法性能明显优于网格搜索法和遗传算法。     

基于人工蜂群算法的支持向量机参数优化及应用

为了解决常用的支持向量机(SVM)参数优化方法在寻优过程不同程度的陷入局部最优解的问题,提出一种基于人工蜂群(ABC)算法的SVM参数优化方法。将SVM的惩罚因子和核函数参数作为食物源位置,分类正确率作为适应度,利用ABC算法寻找适应度最高的食物源位置。利用4个标准数据集,将其与遗传(GA)算法、蚁群(ACO)算法、标准粒子群(PSO)算法优化的SVM进行性能比较,结果表明,本文方法能克服局部最优解,获得更高的分类正确率,并在小数目分类问题上有效降低运行时间。将本文方法运用到计算机笔迹鉴别,对提取的笔迹特征进行分类,与GA算法、ACO算法、PSO算法优化的SVM相比,得到了更高的分类正确率。     

一种新的多类模式识别方法

提出了一种基于支撑向量机的多类分类器,用N-1个支撑向量机组合构成一个具有二叉树结构形式的N-多类分类器．讨论了该多类分类器的泛化推广能力,同时还提出了该多类分类器的基于特征空间的BTSVM学习算法,BTSVM算法使用核函数转换的方式计算特征空间的样本距离;采用类间最小距离最大化作为聚类准则,在每个决策结点产生两个最优子集;然后采用支撑向量机学习算法学习两个最优子集,确定决策结点的最优分类面．理论和实验结果表明,本文提出的基于支撑向量机的多类分类器在整体性能上要优于其它类似的分类器系统。     

基于混合SVM与AdaBoost分类的面部表情识别的人机交互

基于二维主元分析（2DPCA）方法提出一种混合支持向量机（SVM）与AdaBoost算法的面部表情分类方法。首先,该方法对灰度图像进行人脸检测,通过小波变换和二维主元分析得到特征数据,有效地减少了计算量;然后,采用SVM方法对特征数据进行分类学习,得到初始分类器;最后,通过AdaBoost算法对SVM分类结果进行进一步加强,形成强分类器,提升了分类能力,确保了表情识别工作,并实现基于面部表情识别的智能轮椅的人机交互的鲁棒性。实验结果表明：该方法不仅有效地提高了样本的分类能力,而且降低了计算的复杂度,在智能轮椅人机交互实验中的平均识别率达到92.5%。     

Color classification using margin-setting with ellipsoids

This paper presents a robust and computationally efficient genetic algorithm for color classification. It designs well-fitted color space prolate spheroids (ellipsoids) that envelop the training pixels. The ellipsoids are then used to classify unlabeled image pixels in accordance with their color, in order to partition the image. The color classification algorithm described here has very low error rates, boasts very high operational speed, and permits trading higher indecision rates for lower rates of misclassification. The performance of the color classifier developed in this paper is compared with those of the support vector machine (SVM) and the nearest-neighbor (kNN) classifiers. It has been shown that our color classifier outperforms SVM and kNN for partitioning of color images that contain several closely spaced color classes. It has higher correct classification, lower misclassification, and significantly reduced operational latency in comparison with color classifiers based on kNN and SVM.     

A study on several machine-learning methods for classification of malignant and benign clustered microcalcifications

In this paper, we investigate several state-of-the-art machine-learning methods for automated classification of clustered microcalcifications (MCs). The classifier is part of a computer-aided diagnosis (CADx) scheme that is aimed to assisting radiologists in making more accurate diagnoses of breast cancer on mammograms. The methods we considered were: support vector machine (SVM), kernel Fisher discriminant (KFD), relevance vector machine (RVM), and committee machines (ensemble averaging and AdaBoost), of which most have been developed recently in statistical learning theory. We formulated differentiation of malignant from benign MCs as a supervised learning problem, and applied these learning methods to develop the classification algorithm. As input, these methods used image features automatically extracted from clustered MCs. We tested these methods using a database of 697 clinical mammograms from 386 cases, which included a wide spectrum of difficult-to-classify cases. We analyzed the distribution of the cases in this database using the multidimensional scaling technique, which reveals that in the feature space the malignant cases are not trivially separable from the benign ones. We used receiver operating characteristic (ROC) analysis to evaluate and to compare classification performance by the different methods. In addition, we also investigated how to combine information from multiple-view mammograms of the same case so that the best decision can be made by a classifier. In our experiments, the kernel-based methods (i.e., SVM, KFD, and RVM) yielded the best performance (Az = 0.85, SVM), significantly outperforming a well-established, clinically-proven CADx approach that is based on neural network (Az = 0.80).     

Relevance vector machine for automatic detection of clustered microcalcifications

Clustered microcalcifications (MC) in mammograms can be an important early sign of breast cancer in women. Their accurate detection is important in computer-aided detection (CADe). In this paper, we propose the use of a recently developed machine-learning technique--relevance vector machine (RVM)--for detection of MCs in digital mammograms. RVM is based on Bayesian estimation theory, of which a distinctive feature is that it can yield a sparse decision function that is defined by only a very small number of so-called relevance vectors. By exploiting this sparse property of the RVM, we develop computerized detection algorithms that are not only accurate but also computationally efficient for MC detection in mammograms. We formulate MC detection as a supervised-learning problem, and apply RVM as a classifier to determine at each location in the mammogram if an MC object is present or not. To increase the computation speed further, we develop a two-stage classification network, in which a computationally much simpler linear RVM classifier is applied first to quickly eliminate the overwhelming majority, non-MC pixels in a mammogram from any further consideration. The proposed method is evaluated using a database of 141 clinical mammograms (all containing MCs), and compared with a well-tested support vector machine (SVM) classifier. The detection performance is evaluated using free-response receiver operating characteristic (FROC) curves. It is demonstrated in our experiments that the RVM classifier could greatly reduce the computational complexity of the SVM while maintaining its best detection accuracy. In particular, the two-stage RVM approach could reduce the detection time from 250 s for SVM to 7.26 s for a mammogram (nearly 35-fold reduction). Thus, the proposed RVM classifier is more advantageous for real-time processing of MC clusters in mammograms.     

An improved naive Bayesian classifier technique coupled with anovel input solution method [rainfall prediction]

Data mining is the study of how to determine underlying patterns in the data to help make optimal decisions on computers when the database involved is voluminous, hard to characterize accurately and constantly changing. It deploys techniques based on machine learning alongside more conventional methods. These techniques can generate decision or prediction models based on actual historical data. Therefore, they represent true evidence-based decision support. Rainfall prediction is a good problem to solve by data mining techniques. This paper proposes an improved naive Bayes classifier (INCB) technique and explores the use of genetic algorithms (GAs) for the selection of a subset of input features in classification problems. It then carries out a comparison with several other techniques. It compares the following algorithms on real meteorological data in Hong Kong: (1) genetic algorithms with average classification or general classification (GA-AC and GA-C), (2) C4.5 with pruning, and (3) INBC with relative frequency or initial probability density (INBC-RF and INBC-IPD). Two simple schemes are proposed to construct a suitable data set for improving their performance. Scheme I uses all the basic input parameters for rainfall prediction. Scheme II uses the optimal subset of input variables which are selected by a GA. The results show that, among the methods we compared, INBC achieved about a 90% accuracy rate on the rain/no-rain classification problems. This method also attained reasonable performance on rainfall prediction with three-level depth and five-level depth, which are around 65%-70%