一种不精确数据的聚类挖掘方法

在聚类过程中考虑到数据的非确定性，提出了一种改进的K-平均算法——FK-算法。FK-算法思想是减小总均方误差的期望值E(SSE),需特别说明的是对数据对象xi 采用在非确定区域内用非确定密度概率函数pdf f(xi)进行描述。用FK-算法对非确定运动模式的运动对象进行了分析，实验表明考虑数据的非确定因素，在聚类分析处理时有比较精确的结果。

Algorithm in clustering location data for uncertain data mining

To consider data uncertainty in the clustering process, this paper proposed a FK-means clustering algorithm that enhanced the K-means algorithm to the goal of minimizing the expected sum of squared errors E(SSE). Specially noted that a data object xi was specified by an uncertainty region with an uncertainty pdf f(xi). This paper applied FK-means to the particular pattern of moving-object uncertainty. Experimental results show that by considering uncertainty, the clustering algorithm can produce more accurate results.