摘要
针对传统 ＦＣＭ聚类算法的不足，提出了具体改进的方法．通过对算法中聚类数Ｃ的选取、隶属度的修正等问题的讨论，有效地弥补了传统算法中存在的不足，取得了较理想的聚类效果．最后通过计算机仿真给出了两种算法的实验结果．
In this paper, an approach is proposed to improve an FCM clustering algorithm. A few ques- tions are discussed, such as how to select clustering number C and how to correct the membership degree, by which, the shortcoming of the conventional algorithm can be overcome and better clustering effect gained. Finally, experimental results of two algorithms are worked out by computer simulation.
出处
《空军雷达学院学报》
2001年第4期27-29,共3页
Journal of Air Force Radar Academy
