摘要
主成分分析(Principal Component Analysis,PCA)可以处理大量过程参数间的关系与变化,排除次要因素,提取主要因素。文章将主成分分析和决策树C4.5算法相结合,提出一种心脏病诊断预测的新方法,该方法采用PCA方法对影响心脏病诊断的众多变量进行降维处理,减少了预测模型的输入量,消除输入数据间的相关性,用C4.5算法建立心脏病诊断的预测模型。经实验证明有效的提高了C4.5算法的分类正确率,提高了心脏病诊断的正确率。
Principal Component Analysis （PEA） can handle a large number of process parameters and changes the relationship between the exclusion of secondary factors, extraction of the main factors. The combination of the principal component analysis and decision tree algorithm C4.5, has been used in the heart disease diagnosis. The new forecast method first use PeA for data dimensionality reduction, reducing the input of the prediction model. Then the C4.5 algorithm has used to establish the prediction model of the heart disease diagnosis. The experiment proved that the result of this method was more accurate than the C4.5 algorithm, and it improved the diagnostic accuracy rate of heart disease.
出处
《计算机与数字工程》
2009年第10期171-174,共4页
Computer & Digital Engineering
