基于指数可能性的新聚类算法

Novel Clustering Algorithm Based on Exponential Possibilistic Theory

鉴于选择合理类别数对聚类的重要性,该文在指数可能性模型的基础上,通过一致函数,提出了基于指数可能性的新聚类方法,实验表明该方法能够在一定范围内确定合理类别数的同时实现有效的聚类,并且在图像分割中也能够获得合理的分割效果。

It is highly important for clustering algorithms to choose the rational number of clusters.Therefore,in this paper,rooted on the exponential possibilistic model,a novel clustering method based on exponential possibilistic theory is presented with the consistent function.The experimental results demonstrate that the new method can ensure choosing the reasonable number of clusters in a dataset within certain range and realizing clustering very well simultaneously.Meanwhile,the rational results can be attained by the new method in image segmentations as well.