| 基于支持向量机的诊断方法研究 |
| |
| 引用本文： | 郑小洪,;侯志强,;李冀鑫,;顾新锋. 基于支持向量机的诊断方法研究[J]. 适用技术之窗, 2008, 0(7): 9-11 |
| |
| 作者姓名： | 郑小洪, 侯志强, 李冀鑫, 顾新锋 |
| |
| 作者单位： | [1]海军航空工程学院研究生管理大队,山东烟台264001; [2]海军航空工程学院飞行器工程系,山东烟台264001 |
| |
| 摘 要： | 根据支持向量机原理，对线性可分与线性不可分两种情况分别建立了分类模型，模型的求解转化为一个二次规划问题，在选用径向基核函数的参数时运用网格搜索的方法进行选取最优参数。在应用到乳房肿瘤的医疗诊断中，准确率为93．00%，较以前的方法有了明显的提高。
|
| 关 键 词： | 支持向量机 特征提取 诊断 |
The Diagnosis Based on Support Vector Machine |
| |
| Affiliation: | Zheng Xiaohong Hou Zhiqiang Li Jixing Gu Xinfeng(1.Graduate Students＇ Brigade, Naval Aeronautical Engineering Institute, Shandong Yantai 264001; 2. Department of Airborne Vehicle Engineering, Naval Aeronautical Engineering Institute, ShandongYantai 264001) |
| |
| Abstract: | According to supporting vector machine principle, the classification models were built to linear and unlinear dividing circumstances, the solution was found by changing to a quadratic programming question respectively, and choosing the mOst excellent parameter with the method of grid search at the parameter luck to choose the radial base kernel Function. It can be applied in medical inspection of breast tumor and the veracity is 93.00%, which is better than those of other methods. |
| |
| Keywords: | Support Vector Machine Feature Selection Diagnosis |
| 本文献已被 维普 等数据库收录！ |
|
