中文摘要：

为在数据缺失的情况下进行心脏病诊断并获得较高的准确率,对缺失值进行处理后,利用径向基函数支持向量机,采用交叉验证和网格搜索寻找最佳惩罚参数和关联参数,对UCI Heart数据集进行分类,多分类准确率为81.89%,二分类准确率为89.61%.仿真结果表明,支持向量机网络模型性能稳定,样本追加能力强,训练时间短,分类效果好,在心脏病等医疗诊断中有很大的应用潜力.

英文摘要：

In order to diagnose the heart diseases at a high accuracy with missing data,we use radial basis function support vector machine to find the best penalty parameters and relevant parameters by applying cross validation and grid searching after processing the missing value,heart,we get the accuracy for multi-classification is 81.89%,and 89.61%of biclassification for classifying the UCI Heart dataset.The results of the simulation indicated that the support vector machine network is stable,high ability to add samples,short training time,good effect for classification and has great potential to be applied in the medical diagnosis for diseases such as heart diseases.