基于改进C-均值聚类算法的空中目标分类

空中目标的正确分类对防空作战效果有着重要的影响。在分析了空中目标类型和目标的表征方法之后,构建了动态聚类中的一种分类算法———改进C-均值算法。该算法改变了以前算法中只用一个聚类中心作为一类的代表的做法,而采用的是类核函数代替类心的方法。类核函数较一个聚类中心更能够表现出某一类的分布信息,从而提高分类的正确率。

Air targets classification based on ameliorative C-means clustering algorithm

The validity of air targets classification has an important influence on aerial defence battle.After analyzing aerial targets types and targets token means,ameliorative C - means which belongs to dynamic clustering algorithm is constructed.C -means changes the method of using one clustering core replacing one class, and it adopts one method of using pit function replacing one class. The pit function can bring out more information about one class than one clustering core and the validity of classification is improved.