您的位置:首页  > 论文页面

基于PSO的变形SVM算法参数选优及应用

发表时间:2011-11-30  浏览量:1549  下载量:382
全部作者: 王洁,白静,刘晓峰
作者单位: 太原理工大学信息工程学院;太原理工大学理学院
摘 要: 支持向量机(support vector machine,SVM)中参数的选择直接影响其机器学习的性能,但现有研究中对SVM参数的选择仍无统一方法。为解决SVM中参数难以选择的问题,使用改进的支持向量机v-SVM,并用粒子群优化算法(particle swarm optimization, PSO)对v-SVM算法中的参数进行选优,并将优选后的参数用于语音识别系统,实验结果表明:该方法有效可行,优化后的参数使支持向量机具有良好的泛化能力,且其语音识别结果较一般v-SVM参数选择方法有很大的提高。
关 键 词: 信号与信息处理;v-SVM;粒子群算法;语音识别
Title: Parameters optimization and application of changed SVM based on PSO
Author: WANG Jie, BAI Jing, LIU Xiaofeng
Organization: College of Information Engineering, Taiyuan University of Technology; College of Science, Taiyuan University of Technology
Abstract: The standard support vector machine (SVM) is a common method of machine learning, and the parameters selection of SVM affects the machine learning ability directly, but present researches do not have uniform method. For avoiding the difficult problem of selecting parameters, this paper uses the changed SVM which is v-SVM, and selects parameters of v-SVM based on particle swarm optimization (PSO), and applies the optimized parameters in the speech recognition system. The results of the experiment show that the method is effective and feasible, and the optimized parameters make v-SVM have good generalization. Comparing with the general method the recognition results have improved.
Key words: signal and information processing; v-support vector machine; particle swarm optimization; speech recognition
发表期数: 2011年11月第22期
引用格式: 王洁,白静,刘晓峰. 基于PSO的变形SVM算法参数选优及应用[J]. 中国科技论文在线精品论文,2011,4(22):2055-2060.
 
0 评论数 0
暂无评论
友情链接