一种改进的集成算法RSCA在医学领域中的应用  

APPLICATION OF AN IMPROVED INTEGRATED ALGORITHM (RSCA) TO MEDICAL DOMAIN

作　　者：王小凤[1] 耿国华[1] 陈燕[1] 周明全[1] 

机构地区：[1]西北大学信息科学与技术学院,陕西西安710069

出　　处：《计算机应用与软件》2007年第12期24-25,48,共3页Computer Applications and Software

基　　金：国家自然科学基金(60372072);中国博士后科学基金(2003033519)

摘　　要：首次将粗糙集和聚类算法结合起来用于医学领域,取得很好效果。先用粗糙集去除数据集中无用的、干扰性的属性进行约简,然后利用一种改进的基于密度的聚类算法对处理后的数据集进行有效的聚类。从结果可以发现用粗糙集处理后的结果比处理前的分类准确度提高很多。此外,算法的时间复杂度低,适合对大规模数据库的挖掘。最后,通过仿真实验,证明了算法确实具有上述优点。RSCA（ rough set and clustering alogrithm）is integrated and used in medical area for the flint time,which has achieved effective results. In this integrated algorithm, RS（ rough set） is used to reduce useless and interfering attributes, and then an improved CA （clustering algorithm） based on density is applied to cluster the data obtained from RS. It can be found from the results that the accuracy of the clustered data is much better than those with no attributes reduced. Furthermore, the algorithm is of low time complexity, and is suitable for mining databases of large scale. Simulative experiment show that this algorithm can be used to analyze medical data effectively.

关 键 词：粗糙集 聚类 密度 约简 

分 类 号：TP314[自动化与计算机技术—计算机软件与理论]