Research On Fuzzy Clustering Method Based On Pair Constraint

With the rapid development of big data and artificial intelligence,data has also become complex,which has led to people's attention to the diversity of data.The diversity of data has led to the derivation of uncertain data,and people have to pay more attention to uncertain data.The uncertainty of the data mainly includes the uncertainty of the data itself and the level of the data attributes.This work investigated the conclusive and indeterminate data.A new fuzzy C-means clustering algorithm is proposed for determining that the data is randomly selected as the initial cluster center and sensitive to the initial cluster center when determining the data.Moreover,a fuzzy clustering algorithm for processing uncertain data is proposed.As the fuzzy C-means clustering method in the fuzzy clustering algorithm is sensitive to the initial clustering center,this work proposes a new fuzzy clustering algorithm based on the pair-constraint method.Firstly,the algorithm introduces a pairwise constraint approach to determine the relationship between the data in the dataset.Secondly,the data in the original dataset is divided into a number of different initial class clusters based on two corollary of pairwise constraints.At last,the fuzzy entropy method is introduced,and the objective function is improved to obtain a new objective function,and the newly obtained objective function is biased to obtain the cluster center function and the membership function,so as to determine the cluster center point and comple te the clustering of data.These results show that the algorithm in this work has a good clustering effect and accuracy compared with the experimental comparison algorithm.In order to effectively solve the uncertain data of spatial attribute values,based on the traditional fuzzy C-means clustering algorithm,a fuzzy C-means clustering algorithm for uncertain data based on pairwise constraints is proposed.First of all,the similarity of uncertain data objects is calculated by KL-distance,and the original data set is divided into clusters of multiple different categories according to the pairwise constraint method.Secondly,the local density and k-nearest neighbor methods are used to determine the indeterminate data objects,and the data in the cluster are filtered for the initial cluster center.Finally,the expected distance of uncertain data is introduced,and the objective function of the traditional fuzzy C-mean is improved.Furthermore,a new objective function for processing uncertain data is constructed,and the final cluster center is obtained,and the clustering of uncertain data is completed.The results show that the proposed algorithm has a good clustering effect and high accuracy when clustering uncertain data.