SVM中相关参数选择与应用研究

发布时间：2018-06-11 18:38

本文选题：支持向量机 + 高斯核函数　；参考：《山东科技大学》2017年硕士论文

【摘要】：支持向量机是基于统计学习理论框架下提出的一种新型算法,在处理分类问题上具有明显的优势。目前,关于支持向量核函数及其相关参数的选取与研究越来越多,并发展的越来越成熟,但具体怎样选取还没有形成一种统一理论,在日常应用中只能靠经验或对比来选取,本文主要研究的是支持向量机中高斯核参数σ和惩罚参数ζ的选择与应用。鉴于此本文工作如下:首先,介绍了研究的背景、意义、现状及支持向量机中的基本原理与相关知识,为下文做铺垫;其次,介绍了粒子群算法的基本原理,针对支持向量机参数寻优问题,通过分析,提出一种改进粒子群算法,并将该算法应用于高斯核参数及惩罚参数的选取中,通过数值实验与基本粒子群算法、标准粒子群算法进行比较,验证了本文所提出的算法寻找到的参数在分类时间上、准确率上具有较好的优越性;最后,把支持向量机应用到人脸识别分类中。将文中提出的新算法用于高斯核参数及惩罚参数的参数寻优,构造支持向量机高斯核函数分类器,并将该分类器应用到人脸识别中,与基本粒子群算法、标准粒子群算法构造的分类器在错分样本、识别率上等通过实验比较,表明了该算法在分类效果上具有优越性。
[Abstract]:Support vector machine (SVM) is a new algorithm based on statistical learning theory, which has obvious advantages in dealing with classification problems. At present, there are more and more researches on the selection and research of support vector kernel function and its related parameters, and the development is more and more mature. However, how to select the support vector kernel function has not yet formed a unified theory, which can only be selected by experience or contrast in daily application. This paper mainly studies the selection and application of Gao Si kernel parameter 蟽 and penalty parameter 味 in support vector machine. In view of this, the work of this paper is as follows: firstly, the background, significance, current situation, basic principles and related knowledge of support vector machine are introduced, which pave the way for the following; secondly, the basic principle of particle swarm optimization (PSO) is introduced. In order to solve the problem of parameter optimization of support vector machines, an improved particle swarm optimization algorithm is proposed, which is applied to the selection of Gao Si kernel parameters and penalty parameters. Compared with the standard particle swarm optimization algorithm, it is verified that the proposed algorithm has better accuracy in classification time. Finally, support vector machine is applied to face recognition classification. The new algorithm proposed in this paper is used to optimize the parameters of Gao Si kernel and penalty parameters, and the support vector machine (SVM) Gao Si kernel function classifier is constructed, which is applied to face recognition, and to the basic particle swarm optimization (PSO) algorithm. The classifier constructed by standard particle swarm optimization (Swarm Optimization) algorithm is superior in classifying samples and recognition rate through experiments, which shows that the algorithm is superior in classification effect.
【学位授予单位】：山东科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18

【参考文献】