| RBF网络和RBF核支持向量机比较研究 |
| |
| 引用本文： | 朱树先,张仁杰.RBF网络和RBF核支持向量机比较研究[J].计算机仿真,2009,26(5). |
| |
| 作者姓名： | 朱树先 张仁杰 |
| |
| 作者单位： | 上海理工大学光学与电子信息工程学院,上海,200093 |
| |
| 基金项目： | 上海信息产业科技创新资助项目 |
| |
| 摘 要： | RBF网络是模式识别中应用最为广泛的一和神经网络.RBF核函数型支持向量机是一种性能卓越的新型学习机.将这两种学习机进行对比分析,以期在实际应用中做出更好的选择.首先,在理论上分析了这两种学习机在分类原理上的异同.接着,将它们应用于人脸识别,利用ORL人脸图像数据库进行了仿真实验,对比分析它们各自的识别率和泛化能力等性能指标.最后,提出了在应用这两种学习机进行模式识别时应注意的方面.实验结果表明,按照本文提出的两种训练模式,RBF型支持向量机在识别准确率上比RBF网络高出2%到4%.这说明RBF型支持向量机的性能要优于RBF网络.但是RBF网络易于实现,在样本数日足够多的情况下也不不失为一种好的算法.
|
| 关 键 词： | 径向基神经网络 径向基核函数 支持向量机 径向基支持向量机 人脸识别 |
A Comparison of RBF Neural Network with RBF Support Vector Machines |
| |
| ZHU Shu-xian,ZHANG Ben-jie.A Comparison of RBF Neural Network with RBF Support Vector Machines[J].Computer Simulation,2009,26(5). |
| |
| Authors: | ZHU Shu-xian ZHANG Ben-jie |
| |
| Affiliation: | Collage of Optical and Electronic Engineering;University of Shanghai for Science and Technology;Shanghai 200093;China |
| |
| Abstract: | The RBF neural network is widely applied in pattern recognition.RBF SVM is a new type of learning machine with excellent performance.The two learning machines have been compared in this paper,in order to give a better choice when they are used in practical application.First,an analysis of the two learning machines used in pattern recognition is made theoretically.Secondly,the two types of learning machines have been used in human face recognition.Experiments based on ORL face database have been done,and the... |
| |
| Keywords: | |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
