Image Classification Algorithm Based on Support Vector Machine and Fuzzy Membership Function

摘要：

To solve the over-fitting problem in the support vector machine due to the noise and outlier in the samples, in the paper it is proposed that the new Support Vector Machine is based on the close-type membership function through the study of fuzzy membership of fuzzy mathematics degree, double ball membership function and k-nearest neighbor algorithm. The Relative distance between the sample and the class center and the tightness around the sample is in the close-type membership function. Compared with traditional Support Vector Machine i classification results of the Fuzzy Support Vector Machine applied in the image classification is more effective and accurate. The availability of the algorithm is verified in the test.

关键词： Support Vector Machine Fuzzy Support Vector Machine fuzzy membership

作者: Feng Xiufang Chang Jianfeng

作者单位: College of Computer Science and Technology Taiyuan University of Technology Taiyuan 030024 China

会议类型: 国际会议

会议名称: 2010 International Conference on Future Information Technology(2010年未来信息技术国际会议 ICFIT 2010)

会议地点: 长沙

会议语种:英文

页码: 108-112

在线出版日期: 2010-12-14（万方平台首次上网日期，不代表论文的发表时间）