In area like pattern recognition,data mining,machine learning and so on,clustering plays a significant role.However,intuitionistic fuzzy sets (IFSs) and interval-valued intuitionistic fuzzy sets (IVIFSs) cannot describe and process indeterminate and inconsistent information,while single valued neutrosophic sets (SVNSs) and interval neutrosophic sets (INSs) can describe and handle it.So far,the existing clustering techniques scarcely involve SVNSs,and do not involve INSs.Motivated by IFCA and single valued neutrosophic clustering algorithms (SVNCA),the paper firstly proposes another SVNCA using the cosine similarity measure based on a dimension root distance of SVNSs.In the clustering algorithm,we define a dimension root distance measure between SVNSs and its cosine similarity measure between SVNSs,and then present a clustering algorithm based on the cosine similarity measure of SVNSs for clustering single valued neutrosophic data.Then,we extend the clustering algorithm for SVNSs to cluster INSs and propose an interval neutrosophic clustering algorithm (INCA).Then we obtain an illustrative example to show the effectiveness and application of the proposed clustering algorithm under single valued neutrosophic and interval neutrosophic environments.%聚类在数据挖掘、模式识别、机器学习等方面具有重要作用.然而,直觉模糊集(IFSs)和区间直觉模糊集(IVIFSs)无法描述和处理不确定和不一致的信息,而单值中性化的模糊集(SVNSs)和区间神经网络集(INSs)可以描述和处理它.到目前为止,现有的集群技术几乎不涉及SVNSs,也不涉及到INSs.为此基于直觉模糊聚类算法和单值中性聚类算法,首先提出了基于SVNSs维根距离的余弦相似度度量的另一种单值中性聚类算法(SVNCA).在聚类算法中,我们定义了SVNSs与SVNSs之间的余弦相似度度量之阃的维根距离度量,然后提出一种基于SVNSs的余弦相似度度量的聚类算法,用于聚类单值中性化数据.然后,将SVNSs的聚类算法扩展到簇内,并提出了一个区间中性聚类算法(INCA).最后,给出了一个算例,说明了在单值中性化和间隔中性环境下所提出的聚类算法的应用和有效性.
展开▼
