中文摘要：

单病种成本估计是医院成本控制中的重要环节之一。分析了目前单病种成本估计存在的问题，针对单病种成本估计这一多因素、非线性问题，提出了一种基于知识发现（KDD）的单病种成本估计方法。研究了单病种成本估计KDD系统框架和基于医院信息系统（HIS）的单病种成本数据仓库模型；建立了基于粗糙集（RS）和支持向量机（SVM）数据挖掘技术的单病种成本估计模型；最后，以单纯性阑尾炎病例进行了实证研究，该模型获得了较高的预测精度和速度，进而证明了该方法的有效性和优越性，为单病种成本提供了科学有效的估计方法并显著提高了医院成本控制的准确性。

英文摘要：

Single disease cost estimation is one of the most important aspects in hospital cost control. This paper analyzes the existing problems of single disease cost estimation; considering the properties of singledisease cost estimation such as multifactor and nonlinearity, it proposes a method of single disease cost estimation based on KDD （knowledge discovery in database）. First, it develops a framework of KDDsystem dedicated to single disease cost estimation and a data warehouse model based on HIS （hospital information system）. Then it proposes the model of single disease cost estimation based on data miningtechniques-rough sets （RS） and support vector machines （SVM）. An example of cost estimation for pure appendicitis patients is presented, and the model has achieved greater accuracy and. efficiency. The resultverifies the validity and superiority of the method. It provides a scientific and effective estimation method of single disease cost and improves the accuracy of hospital cost control.