研究基于目标函数的模糊聚类算法,并对聚类效果的有效性和参数选择进行了详细分析,在数据挖掘平台中实现该算法,通过设置聚类误差和实时误差两项动态指标来确定最佳的判别方法和参数设置.最后将该算法应用于模型生命表制作的前期分析工作中。
This article describes systematically the Fuzzy C-Means (FCM) Clustering Algorithm based on objective function. The effectiveness and parameter selecting of FCM are analyzed in detail. The algorithm is realized in Data Mining Flat. The best determining method and parameter setting is decided by two dynamic data. Finally the successful application of FCM on preliminary work of Model Life Table is realized.