针对模糊聚类算法在样本特征不明显时不能取得很好的聚类效果 ,以及现有的模糊聚类算法需要事先确定聚类数 ,随机性强、容易陷入局部最优等弱点 ,将核函数和有效性函数引入到模糊聚类中 ,提出了模糊核聚类的自适应算法 .此方法在性能上比经典的聚类算法有了较大的改进 ,取得了更好的聚类效果 .实验结果证实了该方法的有效性和可行性 For the fuzzy clustering algorithm can not get good clustering results when the sample features are not obvious, and the existing fuzzy clustering algorithms need to determine the number of clusters in advance, strong randomness, easy to fall into the local optimal weaknesses, the kernel function and The validity function is introduced into the fuzzy clustering and an adaptive algorithm based on fuzzy kernel clustering is proposed.The performance of the proposed algorithm is improved greatly compared with the classical clustering algorithm and a better clustering result is achieved.Experimental results Confirmed the effectiveness and feasibility of the method