【作者】 周晓丹； 冯少荣； 薛永生；

【机构】 厦门理工学院计算机系； 厦门大学计算机系；

【摘要】 <正>1引言近年来,数据挖掘技术已经得到广泛的应用。数据挖掘的过程包括问题理解与定义、数据采集和预处理、模式抽取、知识解释与评估、知识应用,其中数据预处理阶段占用了80%以上的时间,而缺失数据处理是数据预处理阶段十分重要的一环。传统更多还原

【Abstract】 Dealing with missing data is very important during data mining.At present,the traditional KNN algorithm is one of the most popular methods of dealing with missing data,but if the number of neighbors is few,especially in a situation in which the target is far away from the nearest neighbor,the results of traditional nearest neighbor will not be good.This paper introduces a modified nearest neighbor,which uses simple averages,inverse distance weighted averages and majority votes to synthesize data for artificial source,proposes a new method to choose nearest neighbor so that the proposed approach has the better performance and gets more precise estimated values.更多还原