代价敏感支持向量机在医疗诊断中的应用

Cost Sensitive Support Vector Machines Applied in Medical Treatment

不改变原有的算法,增加一个过程,使原来的分类方法转换为Cost-Sensitive。通过与SVM学习算法相结合,得出一种模型CS-SVM,并将其应用于医学具体实例运用,实验证明,这种方法具有较低的错误分类率,也更能反映实际。

Not change the primary algorithm, only add a process to change the primary algorithm to be Cost-Sensitive. Combining with the SVM approach, get a new learning model CS-SVM, and apply it in medical treatment experiment. The experimental results approve that CS-SVM has lower miss-classification rate, also it can reflect the fact well.